About Us   |   Submission   |   Editorial Team   |   Privacy Statement    |   Current Issue   |   GenAI Policy   |   Templates & Author Guidelines   |   Archives   |   Contacts   |   Journal Updates

 
   

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

AN INVENTORY MODELING FOR DECAYING ITEMS WITH PRICE, STOCK AND RELIABILITY-DEPENDENT DEMAND UNDER MEMORY EFFECTS

 

Narendra Kumar, Liliana Guran, Devendra Kumar, Sanjeet Kumar, Ajay Singh Yadav, Krishan Kumar Yadav

 

A generalized Economic Order Quantity (EOQ) framework is employed to investigate memory effects within an inventory system. Fractional calculus provides an effective mathematical tool for reflecting such memory characteristics in economic model. In this study, a fractional-order inventory model without shortages is developed, where demand is depends on price, on-hand stock, and product reliability, while deterioration is assumed to occur at a constant rate. The developing fractional differential equation is formulated in the Caputo sense, and a memory-kernel approach is used to incorporate dependence on past system states. By applying the Laplace transform technique together with Mittag-Leffler functions, an analytical solution to the model is obtained. The results indicate that the degree of memory in the system can be modulated through the order of the fractional derivative or integral. Furthermore, a sensitivity analysis is performed for both short-memory and long-memory scenarios to determine the key parameters affecting system behavior under varying conditions with the help of MATLAB software.

 

Cite: Narendra Kumar, Liliana Guran, Devendra Kumar, Sanjeet Kumar, Ajay Singh Yadav, Krishan Kumar Yadav AN INVENTORY MODELING FOR DECAYING ITEMS WITH PRICE, STOCK AND RELIABILITY-DEPENDENT DEMAND UNDER MEMORY EFFECTS. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 51-57, DOI: https://doi.org/10.24412/1932-2321-2025-489-51-57

 

RELEVANCE OF INFORMATION AND TECHNICAL SUPPORT OF STAFF IN MAINTENANCE AND REPAIR OF STEAM TURBINE DEVICES OF POWER UNITS OF THERMAL ELECTRIC STATION

 

Farzaliyev Y. Z.

 

The need to improve the efficiency of thermal electric station units increases significantly as their service life increases. If during the standard service life, the efficiency of operation was successfully calculated on the basis of a standard algorithm and experimental energy characteristics, then subsequently the correspondence of the initial energy characteristics to the actual characteristics becomes less and less accurate. In order to ensure the efficiency of power units, it is necessary to reduce the frequency of clarification of empirical characteristics, which causes a significant increase in operating costs. The forced approach to solving operational problems is taking into account operating experience, information about existing defects. In this case, the strict calculation method is replaced by an intuitive approach, the risk of an erroneous solution of which is significant, and improving the method for solving operational problems is transformed from desirable to necessary.

 

Cite: Farzaliyev Y. Z. RELEVANCE OF INFORMATION AND TECHNICAL SUPPORT OF STAFF IN MAINTENANCE AND REPAIR OF STEAM TURBINE DEVICES OF POWER UNITS OF THERMAL ELECTRIC STATION. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 58-68, DOI: https://doi.org/10.24412/1932-2321-2025-489-58-68

 

RELIABILITY ANALYSIS OF A THAP MISSION WITH SEQUENTIAL PHASES

 

Rasm K, Dharmaraja Selvamuthu, Raina Raj, Vladimir Vishnevsky, Dimitry Kozyrev

 

Stratospheric platforms, commonly known as high-altitude platforms (HAPs), are gaining traction as an economical and sustainable option for diverse applications, including environmental monitoring, surveillance, and communication networks. Tethered HAPs (tHAPs), which remain physically connected to the ground, provide improved stability, a continuous power source, and dependable data transfer capabilities. Analyzing the reliability of these systems as phased mission systems (PMS) is essential, given the multiple operational stages they undergo-such as ascend, primary task, and descend, posing distinct operational demands and challenging environmental conditions. Maintaining reliability throughout these phases is crucial for ensuring both mission effectiveness and long-term operational integrity. This research explores the necessity and practicality of conducting reliability assessments on tethered HAP systems by modeling them as PMS with well-defined stages. A comprehensive evaluation is performed for each phase, accounting for potential risks and operational constraints. Furthermore, a numerical example illustrates the reliability assessment methodology, incorporating Markovian techniques to derive key performance indicators and reliability measures for every phase.

 

Cite: Rasm K, Dharmaraja Selvamuthu, Raina Raj, Vladimir Vishnevsky, Dimitry Kozyrev RELIABILITY ANALYSIS OF A THAP MISSION WITH SEQUENTIAL PHASES. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 69-81, DOI: https://doi.org/10.24412/1932-2321-2025-489-69-81

 

A SURVEY ON IMAGE GENERATION TECHNIQUES PARADIGMS, EVOLUTION, DEEP LEARNING ADVANCEMENTS AND FUTURE DIRECTIONS

 

Nrupesh Shah, Dr. Sanjay Patel

 

Image generation techniques have witnessed significant advancements in recent years. Classification of Image Generation Approaches is an important topic in today's rapidly advancing technological landscape. We will examine primary approaches in this field: Generative Adversarial Networks (GANs), Variational Autoencoders (VAEs), Stable Diffusion Processes (SDPs) and other similar approaches. GANs, VAEs and SDPs have shown remarkable performance in terms of image quality, as well as scalability and efficiency. This survey focuses onto seminal works that have shaped the field of image generation, spanning Generative Adversarial Networks (GANs), Variational Autoencoders (VAEs), Autoregressive Models, Flow based image generation as well as diffusion-based approaches, and also compares various evaluation metrices in the field. We also provide a list of challenges in this field.

 

Cite: Nrupesh Shah, Dr. Sanjay Patel A SURVEY ON IMAGE GENERATION TECHNIQUES PARADIGMS, EVOLUTION, DEEP LEARNING ADVANCEMENTS AND FUTURE DIRECTIONS. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 82-104, DOI: https://doi.org/10.24412/1932-2321-2025-489-82-104

 

NEUTROSOPHIC TOPOLOGIZED BIPARTITE GRAPH

 

Balakrishnan A and Thiruganasambandam K

 

Neutrosophic topological bipartite graph analyses unreliability in bipartite structured graphs by incorporating neutrosophic concepts with topologic possessions. This is a novel approach by providing theorems that integrate truth, indeterminacy and falsity values in topology axioms of neutrosophic graphs. The study introduces a topological space where each singleton set is either open or closed, and the boundary of each element is limited to two or fewer connections. And, it extends the topological graph into various graph structures like a neutrosophic star graph, a neutrosophic bistar, a neutrosophic bipartite graph, a neutrosophic tree and a neutrosophic complete bipartite graph through theorems and examples.

 

Cite: Balakrishnan A and Thiruganasambandam K NEUTROSOPHIC TOPOLOGIZED BIPARTITE GRAPH. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 105-113, DOI: https://doi.org/10.24412/1932-2321-2025-489-105-113

 

COMPARATIVE VAR FORECASTING USING LOW AND HIGH-FREQUENCY CONDITIONAL EVT MODELS: EVIDENCE FROM NIFTY INDICES IN INDIA

 

Nadia Sha, Hena M, and Bitha S Mani

 

This study uses Value-at-risk (VaR) and Expected Shortfall (ES) to determine if the availability of high-frequency data improves the accuracy of estimating severe market risk as compared to low-frequency data. The daily closing stock prices of the NIFTY 50, NIFTY 100, and NIFTY 200 Index from January 1, 2021, to April 31, 2025, made up the sample data utilised for analysis. The purpose of the data study was to evaluate the effectiveness of conditional EVT, a two-stage hybrid strategy that merged the EVT methodology with the GARCH, RV, and HAR specification models. Unconditional coverage (UC) and conditional coverage (CC) tests were used to backtest the out-of-sample VaR predictions in order to evaluate their correctness. The regulatory loss function (RLF) and the firm's loss function (FLF) are two utilisation loss functions that were also included in the VaR backtesting process. The generalised breach indicator (GBI) approach was used to backtest the anticipated ES's accuracy. The results of this study demonstrated that when it came to forecasting market risk during times of extraordinary returns, high-frequency conditional EVT that included the HAR specification performed better than low-frequency conditional EVT. When compared to the GARCH-EVT and RV-EVT typed models throughout the periods, the HAR-EVT typed models perform the best according to the VaR and ES measures. The hybrid model of long-memory models for the EVT method is the focus of this research work, which adds to the body of knowledge on the forecasting abilities of risk models.

 

Cite: Nadia Sha, Hena M, and Bitha S Mani COMPARATIVE VAR FORECASTING USING LOW AND HIGH-FREQUENCY CONDITIONAL EVT MODELS: EVIDENCE FROM NIFTY INDICES IN INDIA. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 114-130, DOI: https://doi.org/10.24412/1932-2321-2025-489-114-130

 

EFFICACY OF SINGLE SERVER MARKOVIAN QUEUEING SYSTEM WITH DISSATISFIED CUSTOMERS, BALKING AND NUMEROUS VACATION

 

Iswalikhan.E, Thangaraja.P, Rajeswari.V, Karuppaiya.R, Ramkumar.S, Narmadha.V

 

In this research work, we derive a single server Markovian encouraged arrival queuing system with dissatisfied customers, balking, and numerous vacations, where we examine that encouraged arrival customers are due to the fewer customers, balking, and dissatisfied customers. Once a system-size enhancement is idle, the system server obtains vacation. Suppose the system server detritus is idle, after the sabbatical vacation. We inferred the probability-generating functions (PGF) and got the closed-form interpretation when the system server is numerous. In addition, we obtain other important performance measures and Little's law is also satisfied for this model.

 

Cite: Iswalikhan.E, Thangaraja.P, Rajeswari.V, Karuppaiya.R, Ramkumar.S, Narmadha.V EFFICACY OF SINGLE SERVER MARKOVIAN QUEUEING SYSTEM WITH DISSATISFIED CUSTOMERS, BALKING AND NUMEROUS VACATION. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 131-141, DOI: https://doi.org/10.24412/1932-2321-2025-489-131-141

 

ENHANCING THE ROBUSTNESS OF SUPPORT VECTOR REGRESSION WITH DEPTH-INDUCED WEIGHT FUNCTIONS FOR IMPROVED PREDICTION

 

Kalaivani S

 

Support Vector Regression (SVR) is an extension of Support Vector Machines designed for regression tasks. By using an epsilon-insensitive loss function, SVR balances model complexity and accuracy, making it effective at handling noise and outliers. Data depth functions measure the centrality of data points in multivariate spaces, providing a robust approach to identify influential data based on their position relative to the data distribution. This study introduces Depth-Induced SVR, an enhanced SVR model that incorporates robust depth functions as a weight mechanism with the kernel function to improve efficiency and robustness. Various robust depth functions such as Mahalanobis, Halfspace, L2, Projection, Spatial, and Zonoid are evaluated as weighting functions to determine the most effective combination with the kernel function. Model efficacy is assessed using performance metrics including Mean Squared Error (MSE), Mean Absolute Error (MAE), Mean Absolute Percentage Error (MAPE), and Median Absolute Error (MDAE). This comparative approach aims to identify the optimal depth function for robust SVR performance across diverse real datasets and simulated environments.

 

Cite: Kalaivani S ENHANCING THE ROBUSTNESS OF SUPPORT VECTOR REGRESSION WITH DEPTH-INDUCED WEIGHT FUNCTIONS FOR IMPROVED PREDICTION. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 142-150, DOI: https://doi.org/10.24412/1932-2321-2025-489-142-150

 

MODELING AND IMPROVING THE RELIABILITY OF SMART ELECTRICAL GRIDS

 

Rahila Muradova

 

The transformation of conventional power systems into intelligent electric networks, or Smart Grids, has introduced significant improvements in efficiency, flexibility, and sustainability. However, this evolution also brings new reliability challenges due to complex cyber-physical interactions, integration of distributed energy resources, and increased exposure to cyber threats. This paper presents a comprehensive methodology for assessing and improving the reliability of smart grids through hybrid modeling, real-time analytics, and fault-tolerant system design. Key components with the highest impact on system reliability—such as communication modules and cyber interfaces—are identified and analyzed. A hybrid reliability model combining probabilistic techniques with machine learning methods is developed and illustrated through a case study involving redundant control systems. The results demonstrate that strategic redundancy and predictive diagnostics can significantly enhance the resilience of modern power infrastructures. The study concludes with recommendations for integrating reliability considerations into autonomous grid operation.

 

Cite: Rahila Muradova MODELING AND IMPROVING THE RELIABILITY OF SMART ELECTRICAL GRIDS. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 151-156, DOI: https://doi.org/10.24412/1932-2321-2025-489-151-156

 

MULTIOBJECTIVE NURSE SCHEDULING PROBLEM IN A FUZZY ENVIRONMENT

 

Ahmad Yusuf Adhami, Nabil Ahmed Khan, Ahteshamul Haq

 

Nurse scheduling is vital to hospital quality in almost every aspect, including patient care, bedside assistance, medication administration, support during major surgeries, data collection, reporting, and more. Nurse staffing and recruitment have become critical in the healthcare sector, particularly for staff scheduling, which is essential across industries and services. However, maintaining sufficient nurses is crucial, as understaffing can lead to life-threatening situations with serious financial, social, and legal consequences. This research article aims to minimize scheduling costs by optimizing nurse allocation between departments and shift-wise cost distribution. We assume that the costs associated with various hospital departments are parabolic fuzzy. This study adopts a case study approach to explore the issue of nurse scheduling in a hospital environment, where cost minimization and resource allocation are essential. Fuzzy programming is used to achieve a compromise solution that balances cost efficiency with the required staffing level in hospitals.

 

Cite: Ahmad Yusuf Adhami, Nabil Ahmed Khan, Ahteshamul Haq MULTIOBJECTIVE NURSE SCHEDULING PROBLEM IN A FUZZY ENVIRONMENT. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 157-171, DOI: https://doi.org/10.24412/1932-2321-2025-489-157-171

  

PREDICTIVE MODELING OF POLYCYSTIC OVARY SYNDROME USING MACHINE LEARNING ALGORITHMS

 

Gneya Pandya, Dhara Solanki, Dr. Jigna Jadav, Dr. Kajal Patel

 

PCOS is a prevalent health condition affecting women globally, and if not diagnosed early, it can lead to severe complications like type two diabetes and gestational diabetes. Our research is focused on enhancing the accuracy and reliability of PCOS diagnosis using machine learning techniques. We utilized a public dataset and applied a range of machine learning models like Logistic Regression (LR), Support Vector Machine (SVM), Decision Tree (DT), and Random Forest (RF) ensemble algorithms. By implementing feature selection methods, we aim to identify the most critical factors contributing to the diagnosis, ensuring that the best-performing model is selected for reliable results. Notably, our research achieved a very good accuracy by using the Random Forest algorithm, highlighting its potential to provide a more precise and trustworthy diagnosis of PCOS. This work represents a significant step toward improving healthcare outcomes for women with PCOS.

 

Cite: Gneya Pandya, Dhara Solanki, Dr. Jigna Jadav, Dr. Kajal Patel PREDICTIVE MODELING OF POLYCYSTIC OVARY SYNDROME USING MACHINE LEARNING ALGORITHMS. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 172-179, DOI: https://doi.org/10.24412/1932-2321-2025-489-172-179

 

LOG EXPONENTIATED KUMARASWAMY DISTRIBUTION WITH REAL LIFE APPLICATIONS

 

Mahvish Jan, S.P. Ahmad

 

The generalization of probability distributions plays a crucial role in broadening their applications across various domains. To achieve this, this paper introduces a novel three-parameter generalization of the Kumaraswamy distribution. The new distribution is termed as the Log Exponentiated Kumaraswamy (LEK) distribution. Some essential properties of this distribution are studied and the estimate of parameters are obtained using the maximum likelihood estimation procedure. Also, a simulation study is conducted and two real life data sets demonstrate the proposed distribution's applicability. The performance of the proposed LEK distribution proved to be better than comparative models.

 

Cite: Mahvish Jan, S.P. Ahmad LOG EXPONENTIATED KUMARASWAMY DISTRIBUTION WITH REAL LIFE APPLICATIONS. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 180-190, DOI: https://doi.org/10.24412/1932-2321-2025-489-180-190

 

ANALYSIS OF A SINGLE-SERVER QUEUE WITH ENCOURAGED ARRIVALS, IMPATIENT CUSTOMERS AND DYNAMIC SERVICE SWITCHING

 

Ismailkhan.E, Nagarajan.G, Sundar Raj.M, Immaculate.S

 

This study examines a single-server queueing system that features encouraged arrivals, customer impatience and a self-switching service mechanism. In this model, the arrival rate increases with the number of customers present in the system, specifically modeled as λ(1+η), where λ is a positive constant and η represents the current number of customers.

The server initially operates at a service rate μ1 as long as the system size is below a predefined threshold K. Once the number of customers reaches K, the server switches to a faster service rate μ2. Customers become impatient after entering the queue and may abandon the system at a rate γ. Still, this impatience applies only while the server is operating at the slower service rate μ1. No reneging occurs once the server switches to the faster rate μ2. The steady-state probability distribution of the system is derived and performance metrics such as the expected waiting time and the average number of customers in the system are evaluated. Numerical illustrations are also provided to demonstrate the impact of encouraged arrivals and server switching behavior.

 

Cite: Ismailkhan.E, Nagarajan.G, Sundar Raj.M, Immaculate.S ANALYSIS OF A SINGLE-SERVER QUEUE WITH ENCOURAGED ARRIVALS, IMPATIENT CUSTOMERS AND DYNAMIC SERVICE SWITCHING. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 191-199, DOI: https://doi.org/10.24412/1932-2321-2025-489-191-199

 

A STATISTICAL ANALYSIS OF FUZZY CONFOUNDING FACTORIAL EXPERIMENTS

 

Sri Devi, S. N. Manoharan, T. Karthikeyan, D. Pachiyappan

 

The factorial experiment is used to test more than one factor at a time, whereas simple experiments can test only one factor at a time. The technique of reducing the block size by making one or more interaction contrast identical to block contrast is known as confounding. Only the higher-order interactions are confounded in factorial experiments, and the interaction effects have been confounded within the blocks. This article proposes that confounding design is performed with Latin squares with trapezoidal fuzzy numbers (TrFNs). In order to obtain an unbiased estimate of an error from a Latin square, the rows and columns must be rearranged in random order. Finally, we provide numerical example to clarify the discussions.

 

Cite: Sri Devi, S. N. Manoharan, T. Karthikeyan, D. Pachiyappan A STATISTICAL ANALYSIS OF FUZZY CONFOUNDING FACTORIAL EXPERIMENTS. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 200-208, DOI: https://doi.org/10.24412/1932-2321-2025-489-200-208

 

INVESTIGATION OF THE APPLICATION OF THE NEW INDUCTION SYSTEM IN VERTICAL AXIS WIND GENERATORS

 

I.M. Marufov, N.M. Piriyeva, E.S. Safiyev

 

In recent scientific and technical literature, vertical axis wind generators based on magnetic levitation have demonstrated several advantages over traditional horizontal-axis wind turbines. However, despite these benefits, challenges remain in maximizing the output power, improving the overall performance, and ensuring the long-term reliability and stability of magnetically levitated wind generators. Existing levitation systems need further development to fully realize the potential of vertical axis designs. To address these challenges, a new controllable induction levitator was developed specifically to enhance magnetically levitated vertical axis wind generators. A detailed mathematical model of the controlled induction levitator was established, and the main physical dimensions were determined. The modeling approach employed principles of dimensional compatibility, dimensionless parameters, and generalized mathematical expressions to simplify the solution process. Based on this model, a methodology for parameter optimization was developed, alongside calculation procedures for key design variables. The mathematical model yielded explicit expressions for the levitation constant, the working air gap thickness, and the primary dimensions of the magnetic coil and its windings. These findings provide critical insights for optimizing levitator design and enhancing the performance, reliability, and stability of magnetic levitated vertical axis wind generators. The study confirms that the development of a controllable induction levitation system is essential for advancing vertical axis wind generator technology.

The proposed model and optimization recommendations offer a robust foundation for improving levitation efficiency, increasing power output, and ensuring both operational stability and structural reliability. Further refinement and implementation of control induction supports are expected to significantly enhance the operational capabilities of magnetically levitated vertical axis wind generators.

 

Cite: I.M. Marufov, N.M. Piriyeva, E.S. Safiyev INVESTIGATION OF THE APPLICATION OF THE NEW INDUCTION SYSTEM IN VERTICAL AXIS WIND GENERATORS. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 209-215, DOI: https://doi.org/10.24412/1932-2321-2025-489-209-215

 

SINE LOG-LOGISTIC DISTRIBUTION FOR MODELING REMISSION AND SURVIVAL TIMES OF CANCER PATIENTS

 

Santanu Dutta, Aditya Kumar Yadav

 

Modeling the remission times of bladder cancer patients reported by Lee and Wang [26], the survival times following radiotherapy and chemotherapy of head and neck cancer patients and bone cancer patients reported by Efron [10] and Mansour et al. [16] have attracted considerable research interest. In this paper, we propose a new two-parameter distribution that fits these data well and compares well in model selection for these data with existing distributions that use three or more parameters. Its distribution, density, quantile, and hazard functions have closed-form expressions. The survival function of the proposed model exhibits a regularly varying tail, and its hazard function is either strictly decreasing or inverse-bathtub shaped, depending on the value of the shape parameter. We derive several properties of the proposed distribution, estimate its parameters, and demonstrate its effectiveness in modeling the hazard function and uncovering key features of the remission time and survival time data.

 

Cite: Santanu Dutta, Aditya Kumar Yadav SINE LOG-LOGISTIC DISTRIBUTION FOR MODELING REMISSION AND SURVIVAL TIMES OF CANCER PATIENTS. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 216-237, DOI: https://doi.org/10.24412/1932-2321-2025-489-216-237

 

MANAGEMENT OF ORGANIZATIONAL RISKS IN ENTERPRISES OF THE RUSSIAN OIL AND GAS SECTORBASED ON EXPERT ASSESSMENT RESULTS

 

Evgeny Gvozdev

 

For effective management of integrated safety at enterprises within the segment of the Russian oil and gas sector (ROGS), it has become a tradition to create an integrated safety system (ISS) designed to prevent conditions for the emergence of various risks (technical, organizational-technical, and organizational). The risk-oriented approach used at enterprises all-round has gained widespread recognition over the past decades, and any risk is considered a cause that transforms into a hazardous event (HE), such as an accident or fire. Currently, there exists a variety of methods for risk assessment. The article presents a justified approach for selecting a method to assess organizational risks in ISS, which will allow evaluating the activities of the security directions within its structure, identifying control points with significant shortcomings, forming measures to correct identified deviations, and observing the dynamics of improvement in impact indicators over the analyzed time interval. Aim and objectives. The purpose of the article is to justify a method for assessing organizational risks that have arisen due to deficiencies of personnel in directions implementing control measures concerning the personnel of production structural units. A scientific task has been solved in which the preference for using the expert assessment method, known as the priority ranking method, is justified. It has advantages over other methods. The use of the normal distribution functional together with this method allows experts to select an individual security direction in which organizational risk has arisen. A scientific task, in which an example demonstrating the impact indicators of security personnel directions on the overall state of an overall condition of ISS, has been solved. The possibility of using the priority ranking method in practice and obtaining results in quantitative value (measure) for organizational risks is justified.

 

Cite: Evgeny Gvozdev MANAGEMENT OF ORGANIZATIONAL RISKS IN ENTERPRISES OF THE RUSSIAN OIL AND GAS SECTORBASED ON EXPERT ASSESSMENT RESULTS. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 238-253, DOI: https://doi.org/10.24412/1932-2321-2025-489-238-253

 

OPTIMIZATION AND ANALYSIS OF SKIP-LOT SAMPLING STRATEGIES BASED ON A DOUBLE-SAMPLING FRAMEWORK

 

Seethalakshmi Ramaswamy, Jayaraj. R, D. Kanagajothi

 

This paper presents a novel approach to designing skip-lot sampling inspection plans aimed at reducing the required sample size while maintaining effective quality control. By using the traditional double-sampling plan as a reference, the proposed design provides a framework for making inspection more efficient without compromising the reliability of decisions regarding lot acceptance. The study explores the operational characteristics of the suggested plan and evaluates its performance relative to conventional double-sampling procedures, demonstrating that it can achieve equivalent or better-quality assurance with fewer samples. One of the key advantages of the proposed plan is its lower acceptance rates, which enhances the stringency of quality control while potentially reducing inspection costs and effort. The design process incorporates critical quality metrics, including the acceptable quality level and the limiting quality level as well as the acceptable quality level and the average outgoing quality level to guide the selection of the most appropriate sampling plan for a given production scenario. Overall, the proposed skip-lot inspection approach offers a practical, cost-effective alternative to standard sampling techniques, providing manufacturers and quality managers with a method to optimize inspection resources while ensuring high product quality.

 

Cite: Seethalakshmi Ramaswamy, Jayaraj. R, D. Kanagajothi OPTIMIZATION AND ANALYSIS OF SKIP-LOT SAMPLING STRATEGIES BASED ON A DOUBLE-SAMPLING FRAMEWORK. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 254-261, DOI: https://doi.org/10.24412/1932-2321-2025-489-254-261

 

SENSITIVITY ANALYSIS OF TRIANGULAR AND SYMMETRIC SPLITTING METHODS FOR POSITIVE DEFINITE LINEAR SYSTEMS - BLOCK STOCHASTIC MATRICES

 

Boilepla Harika, Dasari Rajaiah, Renikunta Ramesh, L.P. Rajkumar, and Perati Malla Reddy

 

In this paper, triangular and symmetric(TS) splitting method is applied to regularized linear system of block stochastic coefficient matrix for finding the steady state probability vector and discussed the sensitivity analysis. The homogeneous system is transformed into regularized non-homogeneous linear system by using preconditioned matrix with the small perturbation. The sensitivity analysis depends on the perturbation parameter. From the numerical results, it is concluded that, the sensitivity analysis and convergence analysis of TS method changes rapidly when a small change in the perturbation parameter. Moreover, the numerical value of spectral radius gives the convergence analysis and the bounds of TS method.

 

Cite: Boilepla Harika, Dasari Rajaiah, Renikunta Ramesh, L.P. Rajkumar, and Perati Malla Reddy SENSITIVITY ANALYSIS OF TRIANGULAR AND SYMMETRIC SPLITTING METHODS FOR POSITIVE DEFINITE LINEAR SYSTEMS - BLOCK STOCHASTIC MATRICES. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 262-276, DOI: https://doi.org/10.24412/1932-2321-2025-489-262-276 

 

A STATISTICAL EXPLORATION OF THE EFFECTS OF ADDITIONAL PARAMETERS IN MATHEMATICS

 

G. Aruna, J. Jesintha Rosline

 

The Bipolar Fuzzy Set is an advancement of the classical fuzzy set, developed to represent situations involving both positive and negative preferences or opinions simultaneously. The bipolar score function plays a vital role in decision-making theory. In this article, a bipolar score function is applied to real-life data to examine the effect of additional parameters in the field of mathematics. Furthermore, a research hypothesis is formulated, and the bipolar score function values are statistically analyzed using SPSS software to determine the impact of these additional parameters in mathematics.

 

Cite: G. Aruna, J. Jesintha Rosline A STATISTICAL EXPLORATION OF THE EFFECTS OF ADDITIONAL PARAMETERS IN MATHEMATICS. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 277-284, DOI: https://doi.org/10.24412/1932-2321-2025-489-277-284 

 

RELIABILITY ESTIMATION OF A SYSTEM BASED ON THE STRESS-STRENGTH MODEL USING A WEIBULL DISTRIBUTION

 

K Sruthi and M Kumar

 

Reliability analysis is increasingly used in practical applications. The Weibull distribution is versatile, offering increasing, decreasing, bathtub, and unimodal hazard rates. Its flexibility enables a more accurate representation of real-world data, leading to reliability expressions that are both more realistic and more reliable. In this paper, we analyze the reliability of a single component system using a stress strength model, where both the system strength and stress follow a Weibull distribution. We extend the analysis to situations where a single-component system is exposed to two independent stresses, with the strength also following a Weibull distribution to estimate reliability. Furthermore, we estimate the reliability of a stress-strength model for a single component exposed to n stresses.

Our analysis includes a series system with two independent components under a common stress, using Weibull distributions for stress and exponential distributions for strength. Furthermore, we study the reliability of a series system with n components under common stress, where the stress follows a Weibull distribution. Numerical computations are performed to demonstrate the results. Finally, a comparative study is conducted to evaluate the reliability expression derived in our work using the Weibull distribution against the existing methods.

 

Cite: K Sruthi and M Kumar RELIABILITY ESTIMATION OF A SYSTEM BASED ON THE STRESS-STRENGTH MODEL USING A WEIBULL DISTRIBUTION. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 285-295, DOI: https://doi.org/10.24412/1932-2321-2025-489-285-295

 

ASSESSMENT OF CRUDE PROBABILITIES OF FAILURE FOR TWO-COMPONENT SERIES SYSTEM SHOCK MODEL

 

Saraswati V Katagi and S. B. Munoli

 

A Series system with two components having a single risk on each component is same as a single component with two independent risks acting on it. The third risk may be a catastrophic risk to whole system. A two-component series system shock model with two kinds of shocks namely, damage shock and catastrophic shock, is studied from competing risks perspective. The crude probabilities of different modes of system failure are derived. Information from life testing experiment is used in assessment of crude probabilities for system failure. Results are validated through simulation studies.

 

Cite: Saraswati V Katagi and S. B. Munoli ASSESSMENT OF CRUDE PROBABILITIES OF FAILURE FOR TWO-COMPONENT SERIES SYSTEM SHOCK MODEL. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 296-303, DOI: https://doi.org/10.24412/1932-2321-2025-489-296-303

 

TRANSMUTED X-RAMA DISTRIBUTION: PROPERTIES AND APPLICATIONS

 

Nimmy George, Sindhu E S

 

In this paper, we introduce a new generalized distribution called Transmuted X-Rama distribution (TXRD) by using the quadratic rank transmutation map (QRTM). The probability density function (PDF), cumulative distribution function (CDF), and key moments of the transmuted distribution are meticulously derived to demonstrate how the transmutation parameter influences the shape of the distribution. Additionally, we analyze its reliability function, hazard rate, odds function, entropy and order statistics all of which are crucial in applications involving uncertainty and longevity data. We propose the method of maximum likelihood estimation (MLE) and method of least squares estimation (LSE) for the estimation of parameters of the distribution. We also show how the transmuted distribution may be fitted to real datasets.

 

Cite: Nimmy George, Sindhu E S TRANSMUTED X-RAMA DISTRIBUTION: PROPERTIES AND APPLICATIONS. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 304-317, DOI: https://doi.org/10.24412/1932-2321-2025-489-304-317  

 

AN EFFICIENT SINGLE VARIABLE INVERSION ALGORITHM FOR DERIVING RELIABILITY BOUNDS IN STORAGE SYSTEMS

 

Emad K. Mutar, Zahir A. H. Hassan

 

Ensuring reliability in data survivability schemes for storage systems is challenging due to limited resources. A key part of reliability analysis is selecting a suitable mathematical representation, such as the structure-function. The structure-function maps the overall system state based on the states of its components and can effectively describe systems of various complexities. Many techniques and algorithms depend on identifying all minimal path sets (MPSs) and minimal cut sets (MCSs) to construct the structure and bounds in two-terminal systems. The upper and lower bounds are essential for demonstrating critical states in the system and for the reliability analysis of storage systems. This paper presents three mathematical algorithms for analyzing storage system reliability. The first two algorithms identify MPSs and MCSs using the connection and adjacency matrices of the system’s graph. The third algorithm determines critical states of the reliability function by using the single-variable inversion (SVI) method for the MPSs and MCSs, helping to establish the system’s reliability bounds. The analysis focuses on finding the greatest lower bound (GLB) and the least upper bound (LUB) for reliability. Storage systems are used to demonstrate the computation of MPSs and MCSs and reliability bounds. These algorithms effectively identify the MPSs and MCSs, calculate their bounds, and visually and numerically represent the results.

 

Cite: Emad K. Mutar, Zahir A. H. Hassan AN EFFICIENT SINGLE VARIABLE INVERSION ALGORITHM FOR DERIVING RELIABILITY BOUNDS IN STORAGE SYSTEMS. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 318-332, DOI: https://doi.org/10.24412/1932-2321-2025-489-318-332

 

TERRITORIAL HIERARCHY OF SUSTAINABLE DEVELOPMENT GOALS AND RISK MANAGEMENT

 

Mariia Lesnykh

 

The problem of achieving the sustainable development goals is considered from the perspective of the territorial hierarchy. The necessity of applying a risk-based approach in the implementation of plans for the socio-economic development of territorial entities has been identified. The results of identification and monitoring of key risks of economic development on the example of urban settlements are presented. The interrelation of key risks and significant goals of sustainable development of the research object is revealed.

 

Cite: Mariia Lesnykh TERRITORIAL HIERARCHY OF SUSTAINABLE DEVELOPMENT GOALS AND RISK MANAGEMENT. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 333-338, DOI: https://doi.org/10.24412/1932-2321-2025-489-333-338  

 

RELIABILITY-ORIENTED MICROGRID CONTROL USING REINFORCEMENT LEARNING

 

O. Aliyeva, I. Babazade

 

The growing penetration of renewable energy sources (RES) in modern power systems introduces significant challenges to reliability due to the inherent stochasticity of wind and solar generation. Microgrids, as decentralized energy systems, offer resilience and flexibility but require intelligent control strategies to balance variable generation and demand. This paper proposes a reinforcement learning (RL)-based multi-agent control framework for microgrids, formalized through a Markov Decision Process (MDP) with states reflecting generation, storage, and load variability. Deep RL algorithms—DQN, PPO, and A2C—are compared against conventional methods in a high-fidelity simulation environment with real-world load and RES data. Key reliability indicators, including Loss of Load Probability (LOLP), Mean Time Between Failures (MTBF), and the self-sufficiency ratio, are evaluated. Results demonstrate that PPO achieves a >40% reduction in LOLP compared to rule-based control, ensuring higher resilience and operational adaptability. The study further discusses computational constraints, generalization issues, and cyber-physical vulnerabilities, and proposes solutions based on federated RL, transfer learning, and anomaly detection. The findings underline RL's potential as a cornerstone for reliability-centered energy management in microgrids.

 

Cite: O. Aliyeva, I. Babazade RELIABILITY-ORIENTED MICROGRID CONTROL USING REINFORCEMENT LEARNING. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 339-345, DOI: https://doi.org/10.24412/1932-2321-2025-489-339-345  

 

RELIABILITY ASSESSMENT OF PHASE CHANGE MATERIALS IN TES FOR RENEWABLE ENERGY SYSTEMS

 

G.K. Abdullayeva, A.L. Bakhtiyarov

 

The reliability of phase change materials (PCMs) in thermal energy storage (TES) systems is crucial for the stable and efficient operation of renewable energy infrastructures. This study presents a comprehensive methodology for assessing PCM reliability under repeated thermal cycling, focusing on degradation mechanisms such as latent heat reduction, phase change temperature shifts, and decreased thermal conductivity. Experimental characterization using techniques like Differential Scanning Calorimetry (DSC) and Thermogravimetric Analysis (TGA) provides data on property evolution over thousands of heating-cooling cycles. Failure thresholds for critical parameters are established to define functional limits. An exponential decay model is proposed to quantify degradation kinetics and predict the effective service life of different PCM types, including paraffins, salt hydrates, and eutectic mixtures. Comparative analysis reveals distinct degradation behaviors and informs material selection for TES applications. The integration of empirical data, failure criteria, and predictive modeling supports reliability-centered design and maintenance planning, ultimately enhancing TES system durability and performance in renewable energy applications.

 

Cite: G.K. Abdullayeva, A.L. Bakhtiyarov RELIABILITY ASSESSMENT OF PHASE CHANGE MATERIALS IN TES FOR RENEWABLE ENERGY SYSTEMS. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 346-352, DOI: https://doi.org/10.24412/1932-2321-2025-489-346-352  

 

ON THE TYPE II HALF-LOGISTIC GENERALIZED INVERSE WEIBULL DISTRIBUTION AND ITS APPLICATIONS

 

Saana Rafi Pathayappurakkal, Jabir Bengala, Bindu Punathumparambath

 

Inverse Weibull distribution is frequently used in the survival and reliability analysis owing to its upside down bathtub type hazard function. Generalized inverse Weibull is the flexible generalization of inverse Weibull distribution which has increasing, decreasing and unimodal hazard rates. This paper introduces a novel extension of generalized inverse Weibull distribution which is developed using new type II half logistic generator. We introduced New type II half logistic generalized inverse Weibull distributions. Mathematical and statistical features of the newly derived distribution were examined. The point estimates of the unknown parameters were computed using maximum likelihood estimation, Weighted Least Square, Anderson-Darling and Cramer-von Mises estimator. We have conducted MCMC simulation using R package to validate the estimation algorithms. Finally, we demonstrated the usefulness of the proposed distribution using two real-world failure time data sets.

 

Cite: Saana Rafi Pathayappurakkal, Jabir Bengala, Bindu Punathumparambath ON THE TYPE II HALF-LOGISTIC GENERALIZED INVERSE WEIBULL DISTRIBUTION AND ITS APPLICATIONS. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 353-369, DOI: https://doi.org/10.24412/1932-2321-2025-489-353-369  

 

AN ATTRIBUTE CONTROL CHART BASED ON GENERALIZED EXPONENTIAL-POISSON DISTRIBUTION UNDER HYBRID CENSORING

 

Gokila B and Sheik Abdullah A

 

This paper presents a novel attribute control chart utilizing a Generalized Exponential- Poisson (GEP) distribution to detect the effect of indeterminacy in manufacturing processes. The proposed chart demonstrates enhanced performance compared to traditional GEP-based control charts, particularly in uncertain environments, by effectively monitoring shifts in process performance and exhibiting lower out-of-control average run lengths (ARL). A case study from the automobile industry further highlights its practical applicability. In addition, the paper explores a hybrid censoring scheme that integrates multiple censoring techniques to address incomplete data in survival and reliability analysis. This approach enhances the accuracy of lifetime estimates, especially in cases where data may be incomplete due to time or resource constraints. A comparative analysis of the proposed chart against traditional control charts underscores its superior performance in real-world applications, especially in situations with uncertain process distributions. The charts parameters are carefully optimized to ensure that the ARL for the in-control process closely aligns with a predefined target. The paper demonstrates the charts efficiency through numerical examples and simulation studies, thereby validating its performance and highlighting its practical advantages.

 

Cite: Gokila B and Sheik Abdullah A AN ATTRIBUTE CONTROL CHART BASED ON GENERALIZED EXPONENTIAL-POISSON DISTRIBUTION UNDER HYBRID CENSORING. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 370-380, DOI: https://doi.org/10.24412/1932-2321-2025-489-370-380

 

XAI-SPCNET: A MULTI-LEVEL EXPLAINABLE AI FRAMEWORK FOR PREDICTIVE CONTROL CHART AUTOMATION IN MANUFACTURING QUALITY MONITORING PROCESS

 

V. Anusha, Lilly George, K.S. Harish, Yuvaraja B, A. Manimaran

 

Timely and accurate detection of process abnormalities in modern manufacturing is needed for sustainable quality assurance so that the process can be rectified quickly. Traditional control charts, although good at illustrating deviations, are poor in prediction and heavily dependent on fixed thresholds, which tend to lead to tardy interventions and high rates of false alarms. Recent developments in artificial intelligence (AI) offer predictive options; nonetheless, their lack of transparency as "black-box" systems erodes users' confidence and influences regulatory compliance difficulties. To bridge this void, Explainable AI (XAI) is ever more critical for Statistical Process Control (SPC), delivering not just precise anomaly prediction but also interpretable, transparent explanations.

This paper presents XAI-SPCNet, a new framework incorporating explainability into predictive SPC to facilitate better manufacturing quality assurance. The architecture involves several expert-informed components: (1) Temporal Causal Decision Trees (TCDT-SPE) using Granger causality for adaptive causal splits that are interpretable; (2) Dynamic Rule Embedding Networks (DREN) combining temporal embeddings with Bayesian rule learning for succinct, context-dependent rules; (3) Contrastive Explanation Generator (CEC-SPE) offering contrastive explanations for violations; (4) Sparse Symbolic Regression (SSR-ED) obtaining sparse mathematical decision boundaries; and (5) Multi-Objective Reinforcement Learner with Explanation Optimization (MORLEO) optimizing accuracy, truthfulness, and chart stability. In synthetic and real-world datasets, XAI-SPCNet presents high prediction accuracy (91%), fast inference, concise rule representation, and greater than 95% explanation coverage, thereby promoting transparent, adaptive, and reliable quality control in manufacturing.

 

Cite: V. Anusha, Lilly George, K.S. Harish, Yuvaraja B, A. Manimaran XAI-SPCNET: A MULTI-LEVEL EXPLAINABLE AI FRAMEWORK FOR PREDICTIVE CONTROL CHART AUTOMATION IN MANUFACTURING QUALITY MONITORING PROCESS. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 381-398, DOI: https://doi.org/10.24412/1932-2321-2025-489-381-398

 

CONSTRUCTION OF EXPONENTIAL LIFETIME BASED ATTRIBUTE CONTROL CHART: A BAYESIAN APPROACH

 

Gunasekaran Munian

 

Statistical Process Control (SPC) is a quality control method that employs statistical methods to investigate, monitor, and improve a process. A control chart is a tool for monitoring process performance that employs visual indicators to detect unusual deviations due to assignable causes. This chart compares the values of a quality characteristic to its corresponding control limits. In the quality control process, the control chart is frequently constructed while ignoring parameter uncertainty. The detection of changes in the parameter(s) within the probability distribution of one or more process-related variables is an important aspect of monitoring. Estimating the parameters is essential since it may affect the control chart's long-term performance in in-control or out-of-control conditions. This article introduces a new attribute control chart utilising a Bayesian approach founded on the Exponential lifetime distribution and the Hybrid censoring technique. A Bayesian framework will be utilised to compute the control chart parameters and the average run length. The parameters for the control chart are determined across various value combinations, and the performance of the newly developed control chart is assessed using the Average Run Length (ARL). Numerical examples are provided to elucidate the proposed control chart, and simulated data are employed to demonstrate its potential applications.

 

Cite: Gunasekaran Munian CONSTRUCTION OF EXPONENTIAL LIFETIME BASED ATTRIBUTE CONTROL CHART: A BAYESIAN APPROACH. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 399-410, DOI: https://doi.org/10.24412/1932-2321-2025-489-399-410  

 

ANFIS-BASED COMPUTING FOR BULK QUEUEING FRAMEWORK WITH LOW-BATCH SERVICE, STANDBY SUPPORT AND VACATION STRATEGIES

 

S. Karpagam, N. Aarthy

 

This study investigates a non-Markovian queueing model involving a single server providing bulk and low-batch services. The model incorporates key operational features such as server breakdown and repair, a standby server and multiple vacations. The service times for the primary server, low-batch service, standby server and vacation are governed by general distributions, whereas breakdown and repair follow exponential distributions. While the primary (bulk) server is under repair, the standby server temporarily takes over service. The probability generating function for queue size at any time is determined, and various system performance metrics are evaluated. Numerical examples are used to demonstrate the analytical findings. Furthermore, the analytical results are validated using the Adaptive Neuro-Fuzzy Inference System (ANFIS), which improves the precision and dependability of the model's predictions.

 

Cite: S. Karpagam, N. Aarthy ANFIS-BASED COMPUTING FOR BULK QUEUEING FRAMEWORK WITH LOW-BATCH SERVICE, STANDBY SUPPORT AND VACATION STRATEGIES. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 411-426, DOI: https://doi.org/10.24412/1932-2321-2025-489-411-426

 

A DIFFERENTIATED RELIABILITY MODEL FOR INDUSTRIAL POWER SUPPLY SYSTEMS CONSIDERING THREE TYPES OF EQUIPMENT FAILURES

 

Sevinc Mirzayeva, Natavan Ismiyeva, Fidan Mustafayeva

 

The reliability of power supply systems is a critical factor influencing the operational continuity and safety of industrial enterprises. Conventional reliability assessments often treat all failures uniformly, neglecting the unique characteristics of failure mechanisms such as short circuits, open circuits, and non-actuation of protection devices. This paper proposes a differentiated reliability evaluation model based on non-stationary failure rate functions that reflect the time-dependent nature of each failure type. Using a piecewise failure rate model across three operational phases—burn-in, steady-state, and aging—the methodology accounts for individual failure flows and their cumulative impact on system reliability.

A numerical case study demonstrates the application of the model to a transformer, showing a more accurate prediction of failure-free probability compared to traditional models. The results confirm that short circuits dominate risk during aging, while non-actuation and open circuits also contribute significantly. The model enhances diagnostic precision, supports targeted maintenance planning, and can be extended to system-level reliability assessment for complex industrial energy infrastructures.

 

Cite: Sevinc Mirzayeva, Natavan Ismiyeva, Fidan Mustafayeva A DIFFERENTIATED RELIABILITY MODEL FOR INDUSTRIAL POWER SUPPLY SYSTEMS CONSIDERING THREE TYPES OF EQUIPMENT FAILURES. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 427-433, DOI: https://doi.org/10.24412/1932-2321-2025-489-427-433  

 

A NEW MODIFIED POWER GENERALIZED WEIBULL DISTRIBUTION: PROPERTIES AND APPLICATION

 

A. M. Rangoli, A. S. Talawar and Vijaya Sorganvi

 

Introduction: Survival analysis has become increasing important for various fields like clinical-trails, public health industrial reliability. A major challenge is accurately modelling of failure time data. Using lifetime distribution, we can accurately model the failure time data and can estimate hazard and survival function. In this study we introduced new modified power generalized Weibull (NMPGW) distribution which exhibits bimodal as well as unimodal density patterns, and also, various patterns of hazard curves like increasing, bathtub and decreasing-increasing-decreasing patterns. Methodology: The proposed NMPGW distribution's various density and hazard patterns were studied. Various statistical properties of the model were derived. Estimation of the parameters were done using maximum likelihood estimation (via Expectation Maximization algorithm) and Bayes technique (via Metropolis Hasting algorithm). Also, standard error and confidence limit were estimated. Comparison of various lifetime distributions were done using information criterion. Results: The model shows good fit for hospital dataset in comparison to various lifetime distributions. Estimated the hazard rate which shows decreasing-increasing-decreasing patterns. In comparison to MLE, Bayes estimation gives lower standard error. Kaplan-Meier survival and NMPGW distribution survival curve shows nearer. Conclusion: The NMPGW distribution introduced in this study offers a versatile tool for modeling different hazard rate patterns. The model's strong performance, validated through real hospital data, suggests it could be a valuable addition to survival analysis, outperforming other modified Weibull models in terms of fit and flexibility.

 

Cite: A. M. Rangoli, A. S. Talawar and Vijaya Sorganvi A NEW MODIFIED POWER GENERALIZED WEIBULL DISTRIBUTION: PROPERTIES AND APPLICATION. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 434-448, DOI: https://doi.org/10.24412/1932-2321-2025-489-434-448 

 

STRESS-STRENGTH RELIABILITY UNDER MULTIVARIATE LOG-NORMAL SETUP AND ITS APPLICATION

 

Anirban Goswami, Babulal Seal

 

In this article it is mainly focused on discussion about estimation of stress-strength reliability (R) under log-normal multivariate setup. We propose a procedure to compute and estimate stress-strength reliability of weighted geometric function when both stress and strength follow a dependent log-normal multivariate distribution with two weightage vectors. We consider the principal component analysis to estimate these vectors. MVUE, MLE and Confidence Intervals of R are obtained. Through simulation studies, their performances are compared using different measures. Finally, we provide a real data analysis.

 

Cite: Anirban Goswami, Babulal Seal STRESS-STRENGTH RELIABILITY UNDER MULTIVARIATE LOG-NORMAL SETUP AND ITS APPLICATION. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 449-470, DOI: https://doi.org/10.24412/1932-2321-2025-489-449-470

 

RELIABILITY-CENTERED INNOVATION PLANNING FOR COMPLEX INDUSTRIAL SYSTEMS

 

Saddat Zeynalova, Rana Mammadova, Leyla Rahimova

 

Ensuring the operational reliability of industrial systems is a critical requirement in the context of technical modernization and increasing system complexity. This paper proposes a reliability-centered methodology for evaluating and enhancing the performance of manufacturing assets during innovation-driven transformation. The approach integrates quantitative reliability metrics, including Mean Time Between Failures (MTBF), Mean Time To Repair (MTTR), and operational availability (Ao), into all phases of technical planning and implementation. Statistical modeling based on exponential and Weibull distributions is applied to failure datasets from mechanical engineering systems to estimate degradation rates and failure probabilities. A case study involving digital twins, CMMS integration, and scenario-based failure modeling demonstrates the effectiveness of predictive maintenance and real-time monitoring tools in reducing system downtime. The proposed framework contributes to the unification of reliability analysis and innovation strategy, offering practical tools for minimizing operational risks and optimizing life cycle costs (LCC) in industrial environments.

 

Cite: Saddat Zeynalova, Rana Mammadova, Leyla Rahimova RELIABILITY-CENTERED INNOVATION PLANNING FOR COMPLEX INDUSTRIAL SYSTEMS. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 471-478, DOI: https://doi.org/10.24412/1932-2321-2025-489-471-478

 

A HYBRID APPROACH FOR DETECTION AND MITIGATION OF ROUTING ATTACKS IN RPL USING DEEP LEARNING

 

Deepak Upadhyay, Hiteishi Diwanji

 

The exponential growth of the Internet of Things (IoT) has led to the massive deployment of resource-constrained smart devices in domains such as smart homes, industrial automation, healthcare, and smart cities. To facilitate communication between these devices, the Routing Protocol for Low-Power and Lossy Networks (RPL) has emerged as a widely adopted IPv6-based routing protocol customized for low-power, lossy environments. This study proposes an Intrusion Detection System (IDS) framework that processes raw IoT traffic data to detect and classify RPL routing attacks using a hybrid approach and takes mitigation steps. The IDS framework integrates multiple models, such as a LSTM enhanced with Multi-Head Attention

for sequential pattern recognition, gradient boosting models such as LightGBM and CatBoost for efficient tabular data classification, and a Multi-Layer Perceptron (MLP) to replicate ensemble knowledge in a lightweight manner. This hybrid IDS demonstrates the efficacy of combining deep learning and machine learning models to detect and mitigate RPL routing attacks, providing a practical and scalable solution for securing IoT networks with highest accuracy of 93.26 % with mobility and 91.96 % without mobility for CNN+LSTM+Attention, LightGBM, CatBoost, MLP algorithms.

 

Cite: Deepak Upadhyay, Hiteishi Diwanji A HYBRID APPROACH FOR DETECTION AND MITIGATION OF ROUTING ATTACKS IN RPL USING DEEP LEARNING. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 479-493, DOI: https://doi.org/10.24412/1932-2321-2025-489-479-493

 

RELIABILITY ASSESSMENT OF THERMAL-ELECTRICAL MODELS FOR PHOTOVOLTAIC SYSTEMS IN MATLAB: ASSUMPTIONS VERSUS EVALUATION

 

Rahim Mammadzada

 

Reliable simulation of photovoltaic (PV) panel behavior requires more than electrical accuracy. It must also reflect the thermal realities of panel operation under real-world conditions. This study investigates the reliability of a Simulink-based thermal electrical model of a Jinko JKM325P-72-V monocrystalline module by comparing simulated temperature profiles with those expected under Standard Test Conditions (STC). The model accounts for multilayer thermal mass and material-specific heat capacities. Heat generation is attributed to both unconverted irradiance and resistive

losses, while thermal dissipation is governed by convective heat transfer. By tuning model parameters based on datasheet specifications and empirical coefficients, the simulation achieves close agreement with reference temperature values under standard conditions. The work highlights how common oversimplifications in PV modeling, such as assuming uniform temperature, can lead to optimistic thermal reliability assessments. The approach emphasizes the importance of verifying modeling assumptions against physical intuition and available data to support more trustworthy simulation-based design decisions for PV systems.

 

Cite: Rahim Mammadzada RELIABILITY ASSESSMENT OF THERMAL-ELECTRICAL MODELS FOR PHOTOVOLTAIC SYSTEMS IN MATLAB: ASSUMPTIONS VERSUS EVALUATION. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 494-503, DOI: https://doi.org/10.24412/1932-2321-2025-489-494-503

 

PERFORMANCE MEASURE OF OPTIMAL RESERVE INVENTORY MODEL BETWEEN TWO MACHINES WITH REFERENCE TO TRUNCATION POINT OF THE REPAIR TIME UNDER FUZZY ENVIRONMENT

 

Jagatheesan R and Ramesh R

 

In inventory control theory, various models have been developed to determine optimal stock levels in practical settings. In sequential production systems, reserve inventory is essential to minimize idle time when upstream machine failures occur. This study assumes that machine repair times are exponentially distributed and satisfy the Setting the Clock Back to Zero (SCBZ) property, with the truncation point itself modeled as a random variable following a mixed exponential distribution. Under these assumptions, an optimal reserve inventory model is derived and analyzed in a fuzzy environment. Uncertainty in system parameters is incorporated using fuzzy numbers, which are defuzzified through the Trisectional Fuzzy Trapezoidal ranking method. Numerical examples validate the applicability of the model in real-life scenarios.

 

Cite: Jagatheesan R and Ramesh R PERFORMANCE MEASURE OF OPTIMAL RESERVE INVENTORY MODEL BETWEEN TWO MACHINES WITH REFERENCE TO TRUNCATION POINT OF THE REPAIR TIME UNDER FUZZY ENVIRONMENT. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 504-512, DOI: https://doi.org/10.24412/1932-2321-2025-489-504-512

 

A TAYLOR SERIES EXPANSION METHOD FOR SOLVING LINEAR FRACTIONAL PROGRAMMING PROBLEMS

 

Poonam Kumari

 

This paper proposes an iterative method to solve Linear Fractional Programming (LFP) problems with inequality constraints. The core idea is to approximate the fractional objective function through a series of linear programming (LP) problems, solved successively using updated feasible points. The algorithm begins by selecting a non-zero feasible point that satisfies all inequality constraints. At this point, the fractional objective function is expanded using a first-order Taylor series, resulting in a linear approximation. This converts the LFP problem into a standard LP problem, which can be solved using common optimization techniques such as the simplex method or the graphical method, depending on the problem's size and complexity. The optimal solution of the LP problem is then used as the new feasible point for the next iteration. In each iteration, the fractional objective function is re-linearized around the current solution, and a new LP problem is formed and solved. This process continues until convergence is achieved, that is, when two successive iterations produce the same or sufficiently similar solutions. The final solution is taken

as the optimal solution to the original LFP problem. The effectiveness and practical utility of the proposed method are demonstrated through numerical examples. Results indicate that the approach is computationally efficient and provides accurate solutions. Compared to traditional transformation-based techniques, the proposed method avoids the introduction of auxiliary variables or complex reformulations, offering a more intuitive and implementable solution framework. Furthermore, the approach exhibits potential for extension to more complex fractional programming models, including multi-objective and equality-constrained formulations.

 

Cite: Poonam Kumari A TAYLOR SERIES EXPANSION METHOD FOR SOLVING LINEAR FRACTIONAL PROGRAMMING PROBLEMS. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 513-527, DOI: https://doi.org/10.24412/1932-2321-2025-489-513-527 

 

DESIGNING AND EVALUATION OF BAYESIAN SKIP LOT SAMPLING PLAN - 2 WITH DOUBLE SAMPLING PLANS AS THE REFERENCE PLAN UNDER GAMMA ZERO-INFLATED POISSON DISTRIBUTION

 

Priyadharshini R and Shalini K

 

Sampling plans are an effective statistical method for assessing and deciding whether to approve or decline a lot based on quality inspection. Skip-lot sampling plans focus on inspecting only a fraction of the submitted lots, significantly reducing costs by minimizing time and effort. Skip-lot sampling is intended for immediate use, and it has been demonstrated that the quality of the submitted lots is consistently high. This article presents designing of Bayesian Skip Lot Sampling Plan – 2 (BSKSP – 2) with Double Sampling Plan (DSP) as the reference plan based on Gamma Zero-inflated Poisson (GZIP) distribution. The Zero-inflated Poisson (ZIP) distribution is suitable for count data with a high frequency of zeros, especially when non-conformities are rare. The Gamma distribution serves as a conjugate prior to the ZIP distribution. The Operating Characteristic (OC) function of skip lot sampling plan is derived and numerical illustrations are provided to illustrate the proposed sampling plan.

 

Cite: Priyadharshini R and Shalini K DESIGNING AND EVALUATION OF BAYESIAN SKIP LOT SAMPLING PLAN - 2 WITH DOUBLE SAMPLING PLANS AS THE REFERENCE PLAN UNDER GAMMA ZERO-INFLATED POISSON DISTRIBUTION. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 528-535, DOI: https://doi.org/10.24412/1932-2321-2025-489-528-535

 

SIX SIGMA BASED CONTROL CHARTS FOR MEAN UNDER TWO-PARAMETER EXPONENTIAL DISTRIBUTION

 

Gomathi K and Ganesan A

 

The statistical process control (SPC) method is widely regarded as the most efficient approach for evaluating production processes, balancing sampling costs with chart performance. Recent research has explored this problem in the context of economic planning for control charts, particularly focusing on adaptive control charts. Traditionally, control chart design relies on a fixed structure to determine key parameters such as sample size, sampling interval, and control limits to meet economic or statistical requirements. However, under the exponential distribution, optimizing control charts based on Six Sigma principles can ensure process stability while minimizing costs. This approach helps reduce either the total expected costs over a finite time horizon or the long-term average expected costs.

 

Cite: Gomathi K and Ganesan A SIX SIGMA BASED CONTROL CHARTS FOR MEAN UNDER TWO-PARAMETER EXPONENTIAL DISTRIBUTION. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 536-544, DOI: https://doi.org/10.24412/1932-2321-2025-489-536-544

 

EXPLORING RECIDIVISM THROUGH BAYESIAN SURVIVAL MODELS: A PARAMETRIC APPROACH WITH RSTAN

 

Fahad Ashraf, Mohammad Parvej, Athar Ali Khan

 

This study adopts a fully Bayesian parametric survival modeling approach to examine recidivism, utilizing the recid dataset from the woodbridge package in R. Leveraging the computational capabilities of RStan, four alternative survival distributions Exponential, Weibull, Lognormal, and Frechet are estimated, incorporating covariates reflecting demographics, criminal history, and post-release conditions. To promote model parsimony and interpretability, covariate selection is guided by variable importance measures derived from a machine learning-enhanced Cox model. Model performance is systematically evaluated using the Leave-One-Out Information Criterion (LOOIC) and the Widely Applicable Information Criterion (WAIC) to ensure robust predictive accuracy. The analysis includes detailed posterior summaries, conditional effect plots to assess covariate influences on time to reoffense, and posterior predictive distribution (PPD) plots for model validation. The findings provide meaningful insights into the timing and risk of reoffending and underscore the value of Bayesian survival models in criminological research.

 

Cite: Fahad Ashraf, Mohammad Parvej, Athar Ali Khan EXPLORING RECIDIVISM THROUGH BAYESIAN SURVIVAL MODELS: A PARAMETRIC APPROACH WITH RSTAN. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 545-561, DOI: https://doi.org/10.24412/1932-2321-2025-489-545-561

 

CONTROL CHARTS WITH SIX SIGMA FOR RANGE UNDER TWO-PARAMETER GAMMA DISTRIBUTION

 

Gomathi K and Ganesan A

 

Quality monitoring of goods and services has been practiced for centuries, whether directly or indirectly. However, the modern approach incorporates quantitative methods rooted in statistical principles for quality control. Statistical quality control (SQC) encompasses a variety of problem-solving techniques used in industrial production, some of which rely on straightforward statistical theories for process control and monitoring. Over time, statistical process control (SPC) has evolved to include a broad range of statistical and optimization methods employed by professionals to enhance quality. In this research paper, we present a comprehensive review and development of a Six Sigma-based control chart for range under the Gamma distribution, examining its various aspects and applications.

 

Cite: Gomathi K and Ganesan A CONTROL CHARTS WITH SIX SIGMA FOR RANGE UNDER TWO-PARAMETER GAMMA DISTRIBUTION. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 562-573, DOI: https://doi.org/10.24412/1932-2321-2025-489-562-573

 

JOINT ADAPTIVE PROGRESSIVE TYPE-II CENSORING WITH DUS-EXTREME VALUE DISTRIBUTION

 

Rakhi Chandran, Chacko V M

 

Statistical estimation of parameters under different censoring schemes is important while using distributions like the Extreme value and Weibull distributions for modelling real lifetime data. Since Extreme value distribution arise as log transformation of the Weibull distribution, there are several advantages for discussing the estimation of extreme value parameters. Moreover, DUS transformation provides more appropriate model without increasing parameters. This study examines point estimation methods, including maximum likelihood and Bayesian approaches, for DUS Extreme Value (DUS-EV) distribution under a joint adaptive progressive Type-II censoring scheme. The proposed scheme offers notable advantages in reducing cost and experimental time. Bayesian estimation is implemented via the Metropolis- Hastings algorithm within a Gibbs sampling framework. Interval estimation is carried out using asymptotic confidence intervals, Highest Posterior Density (HPD) credible intervals, and bootstrap confidence intervals. A comparative analysis of classical and Bayesian results is presented, and the applicability of the proposed methods is demonstrated through real data analysis.

 

Cite: Rakhi Chandran, Chacko V M JOINT ADAPTIVE PROGRESSIVE TYPE-II CENSORING WITH DUS-EXTREME VALUE DISTRIBUTION. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 574-589, DOI: https://doi.org/10.24412/1932-2321-2025-489-574-589

 

DESIGN AND ANALYSIS OF A BAYESIAN ADAPTIVE SKIP-LOT SAMPLING PLAN (BA-SkSP) WITH SINGLE SAMPLING PLAN AS REFERENCE PLAN

 

P. Umanaheswari

 

This research describes a new Bayesian Adaptive Skip-Lot Sampling Plan (BA-SkSP) system designed for quality inspection of constantly manufactured lots using Bayesian statistical concepts. The proposed strategy uses a single sampling plan as a reference plan and includes a Bayesian framework for dynamically adjusting sample frequency depending on prior quality history and real-time inspection outcomes. The construction approach and

performance measures, such as operational characteristic curves, average sample number, average outgoing quality, and average total inspections, are based on Bayesian inference and Markov chain analysis. The performance of the proposed BA-SkSP is examined using extensive simulation studies and compared to traditional SkSP-2 plans under various quality situations. The results show that the Bayesian adaptive technique offers more efficient quality control with fewer inspections.

 

Cite: P. Umanaheswari DESIGN AND ANALYSIS OF A BAYESIAN ADAPTIVE SKIP-LOT SAMPLING PLAN (BA-SkSP) WITH SINGLE SAMPLING PLAN AS REFERENCE PLAN. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 590-598, DOI: https://doi.org/10.24412/1932-2321-2025-489-590-598

 

SOME STATISTICAL PROPERTIES AND APPLICATIONS OF LOMAX GENERATED DISTRIBUTION

 

Brijesh P. Singh, Manisha Ailani & Utpal Dhar Das

 

In this paper, a probability distribution has been proposed using the concept of induced distribution, with the classical Lomax distribution as the baseline. Various statistical properties including moments, entropy measures, reliability functions, and generating functions are derived. The maximum likelihood estimation (MLE) method is used for parameter estimation. Real data application is performed to validate the suitability of the proposed model in comparison to existing distributions.

 

Cite: Brijesh P. Singh, Manisha Ailani & Utpal Dhar Das SOME STATISTICAL PROPERTIES AND APPLICATIONS OF LOMAX GENERATED DISTRIBUTION. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 599-617, DOI: https://doi.org/10.24412/1932-2321-2025-489-599-617

  

OPTIMIZATION OF A DETERIORATING INVENTORY MODEL WITH TIME-DEPENDENT HEAVISIDE DEMAND AND RETURN POLICY UNDER TRIANGULAR FUZZY DECISION-MAKING

 

Kapil Dave, Tanuj Kumar

 

This study aims to develop a two-warehouse inventory model under both crisp and fuzzy environments, incorporating a return policy. In high-demand retail settings such as supermarkets and corporate markets, storage capacity is often limited. When suppliers offer attractive price discounts for bulk purchases or when the item is seasonal such as agricultural harvest outputs-large quantities are procured. However, these quantities may exceed the capacity of the primary warehouse (own warehouse or OW). In such situations, an additional storage facility, referred to as the rented warehouse (RW), is hired. This RW is typically located at a distance and incurs higher holding costs compared to OW. Consequently, inventory is stored in OW first, and only the surplus is placed in RW. To minimize costs, items from RW are transferred continuously to OW following a release pattern. Return policies offered by sellers serve as an incentive for customers. These policies permit buyers to return unsatisfactory products, with a refund typically amounting

 

Cite: Kapil Dave, Tanuj Kumar OPTIMIZATION OF A DETERIORATING INVENTORY MODEL WITH TIME-DEPENDENT HEAVISIDE DEMAND AND RETURN POLICY UNDER TRIANGULAR FUZZY DECISION-MAKING. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 618-634, DOI: https://doi.org/10.24412/1932-2321-2025-489-618-634

 

RELIABILITY RISK PROFILING OF TECHNICAL SYSTEMS USING PARAMETRIC COX REGRESSION WITH WEIBULL BASELINE

 

Iyappan M, G. Sathya Priyanka, Soniya. K, Priyanka. G

 

Reliability analysis of technical systems is vital in predicting failure behavior and improving system design under operational variability. Traditional reliability models often assume constant hazard rates, limiting their ability to model real world degradation mechanisms. This paper presents a Weibull based Cox Proportional Hazards Model to analyse system failure times with covariate influences such as operations stress, environmental temperature, and maintenance interventions. Simulation data is generated for 1000 units under controlled covariate conditions with right censoring. The model estimates hazard functions and survival probabilities using maximum and partial likelihood methods. The results show that the inclusion of time independent covariates significantly enhances the model’s predictive ability for time to failure. The Weibull baseline provides flexibility in representing increasing or decreasing hazard trends typical in engineering systems. This study validates the model’s robustness through simulation and establishes its applicability in predictive maintenance and reliability risk management.

 

Cite: Iyappan M, G. Sathya Priyanka, Soniya. K, Priyanka. G RELIABILITY RISK PROFILING OF TECHNICAL SYSTEMS USING PARAMETRIC COX REGRESSION WITH WEIBULL BASELINE. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 635-641, DOI: https://doi.org/10.24412/1932-2321-2025-489-635-641

 

OPTIMIZING TRAFFIC SIGNAL TIMING FOR EFFICIENT FLOW THROUGH URBAN CROSSROADS

 

Bharathi Ramesh Kumar, S. Geetha, S. Gayathri, A. K. Bhuvaneswari

 

This paper presents an intelligent traffic signal control system aimed at optimizing vehicle flow at a two-lane crossroads intersection using a network flow-based framework. Rapid urbanization and the corresponding rise in vehicle usage have exacerbated congestion issues, especially at intersections where traditional fixed-time signal controls often result in inefficiencies and increased vehicle delays. The proposed system addresses these limitations by integrating real-time traffic data with principles of network flow theory to coordinate signal timing and vehicle movement dynamically. The intersection is modeled as a directed graph, where nodes represent entry and exit points, and edges signify possible vehicle paths. The system utilizes a weighted signal control algorithm that incorporates traffic volume, direction, and flow density to determine optimal signal phases. These weighted parameters guide the adaptive timing of traffic lights, allowing for parallel vehicle movement while preventing collisions. Specific constraints are applied to eliminate conflicting movements and maintain safety, thereby enhancing throughput and reducing total waiting times. A key objective of the model is to minimize total waiting time across all directions by identifying and enforcing necessary conditions for optimal traffic flow. To evaluate the performance of the proposed system, a simulation environment was developed to mimic real-world traffic conditions under various scenarios. Performance metrics such as average delay, queue length, and throughput were used for comparison against traditional signal control methods. Simulation results indicate significant performance improvements. The proposed system achieves a vehicle flow time of 804 seconds compared to 400 seconds in the existing system, with an optimized objective function value of 972. A numerical case study further illustrates the implementation, signal phase structuring, and resulting efficiency gains. In conclusion, the intelligent traffic signal control model demonstrates strong potential for reducing congestion and improving traffic management at urban intersections. Its adaptive, data-driven approach not only enhances current traffic flow but also allows for forecasting future patterns. The model is scalable, flexible, and suitable for integration with intelligent transportation systems, offering a proactive solution to modern traffic challenges.

 

Cite: Bharathi Ramesh Kumar, S. Geetha, S. Gayathri, A. K. Bhuvaneswari OPTIMIZING TRAFFIC SIGNAL TIMING FOR EFFICIENT FLOW THROUGH URBAN CROSSROADS. DOI: https://doi.org/10.24412/1932-2321-2025-489-642-650

 

RELIABILITY AND PERFORMANCE ANALYSIS OF AN UNRELIABLE RETRIAL QUEUEING SYSTEM WITH NEGATIVE ARRIVALS

 

Belaiza Dihia, Taleb Samira

 

In this paper, we focus on a single repairable server retrial queueing system with negative arrivals and two types of breakdowns (passive and active). We assume that the customer in service is lost whenever an active breakdown occurs. Once the repair is completed, the server returns to an idle state. To obtain the system performance metrics, the matrix geometric technique is used combined with direct truncation method. From the standpoint of reliability analysis, we investigate the mean time to the first failure of the system by using the Laplace transform method. Finally, we study numerically the impact of system metrics on the mean number of customers in the orbit, the server state probabilities, the reliability function and the mean time to the first failure.

 

Cite: Belaiza Dihia, Taleb Samira RELIABILITY AND PERFORMANCE ANALYSIS OF AN UNRELIABLE RETRIAL QUEUEING SYSTEM WITH NEGATIVE ARRIVALS. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 651-664, DOI: https://doi.org/10.24412/1932-2321-2025-489-651-664

  

STOCHASTIC PETRI NET ANALYSIS OF RETRIAL REPAIRABLE SYSTEMS WITH TWO REPAIR MODES

 

Kheireddine Boudehane, Samira Taleb

 

This paper presents a dependability analysis of an L-out-of-n system supported by a hierarchical standby structure with warm and cold units. The model accounts for realistic operating conditions, including a single unreliable repair server that may break down but can still provide limited service during failure periods. A retrial mechanism is also considered, where failed units that find the server busy join an orbit and retry for service later. The system is modeled using a Generalized Stochastic Petri Net (GSPN), which is then mapped to a Continuous-Time Markov Chain (CTMC) for steady-state performance evaluation. Furthermore, transient reliability analysis is conducted to derive the survival function of the system and the expectation of the first passage time to failure.

 

Cite: Kheireddine Boudehane, Samira Taleb STOCHASTIC PETRI NET ANALYSIS OF RETRIAL REPAIRABLE SYSTEMS WITH TWO REPAIR MODES. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 665-676, DOI: https://doi.org/10.24412/1932-2321-2025-489-665-676

 

METAHEURISTIC OPTIMIZATION OF RELIABILITY PARAMETERS UNDER VACATION AND INSPECTION POLICIES

 

Diksha Mangla, Shilpa Rani, Shakuntla Singla, Savita Garg

 

This paper explores in detail the reliability and performance evaluation of a repairable system functioning under the combined influence of vacation and inspection policies, using the Regenerative Point Graph Technique (RPGT) as the core analytical framework. The modeling is carried out through a stochastic state-transition representation, where different system conditions such as active operation, failures, repairs, vacation return of the repairman, and inspection mechanisms are explicitly considered. By exploiting the regenerative nature of the process, the study derives closed-form mathematical expressions for key performance indices, including the Mean Time to System Failure (MTSF), steady-state availability, busy period of the server, and the expected number of inspections. These measures provide a structured understanding of how the system behaves and responds under varying operational, repair, and inspection conditions. To further enhance reliability and system performance, the study employs metaheuristic optimization algorithms such as the Particle Swarm Optimization (PSO), the Genetic

Algorithm (GA), and the Cuckoo Search Algorithm (CSA). These optimization methods are implemented to identify the optimal values of system parameters, including failure rate, repair rate, inspection rate, and vacation return rate, which directly influence the efficiency of the system. Numerical experiments and simulation-based illustrations are conducted to validate the theoretical analysis and to compare the effectiveness of the different algorithms. The results demonstrate that CSA and PSO consistently outperform GA, particularly in terms of achieving higher MTSF and availability values, whereas GA, despite slower convergence, provides stable and competitive solutions. The findings emphasize the significance of RPGT as a powerful analytical tool for modeling complex repairable systems and capturing their stochastic dynamics under realistic operational assumptions such as vacations and inspections. Furthermore, the integration of RPGT with evolutionary optimization techniques not only deepens the understanding of system dynamics but also supports practitioners and decision makers in formulating effective reliability and maintenance strategies. This contribution is highly relevant for practical applications in manufacturing industries, communication networks, service organizations, and other industrial systems, where performance optimization and reliability improvement are essential.

 

Cite: Diksha Mangla, Shilpa Rani, Shakuntla Singla, Savita Garg METAHEURISTIC OPTIMIZATION OF RELIABILITY PARAMETERS UNDER VACATION AND INSPECTION POLICIES. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 677-684, DOI: https://doi.org/10.24412/1932-2321-2025-489-677-684

 

IMPROVING RELIABILITY OF LOW-VOLTAGE DISTRIBUTION NETWORKS UNDER LIGHTNING OVERVOLTAGES

 

I.A. Guseynova

 

Low-voltage (LV) distribution networks are highly vulnerable to lightning-induced overvoltages, which represent a significant cause of transformer failures, equipment damage, and service interruptions in regions with high lightning activity. Conventional protective measures—such as improved grounding and primary medium-voltage (MV) surge arresters—are necessary but often insufficient to suppress secondary surges at the LV side. This paper analyzes the physical mechanisms of surge current injection into transformer neutrals, the influence of network topology, and the role of transformer design in generating internal dielectric stress. Comparative evaluation of surge protective devices (SPDs) demonstrates that coordinated deployment of MOV- and hybrid-based arresters at both transformer terminals and consumer service entrances significantly reduces insulation stress and equipment damage. Simulation results and field evidence confirm that such coordinated strategies improve key reliability indicators, including mean time between failures (MTBF) and system average interruption frequency index (SAIFI), while also reducing economic losses. The findings highlight that preventive investments in SPDs are cost-effective, yielding an order-of-magnitude reduction in lightning-related failures. From a reliability-centered asset management perspective, SPD deployment should be considered a strategic measure for enhancing resilience and sustainability of LV distribution networks.

 

Cite: I.A. Guseynova IMPROVING RELIABILITY OF LOW-VOLTAGE DISTRIBUTION NETWORKS UNDER LIGHTNING OVERVOLTAGES. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 685-693, DOI: https://doi.org/10.24412/1932-2321-2025-489-685-693

 

FEASIBILITY AND VIABILITY OF ADAPTIVE RECOMMENDER SYSTEMS: BRIDGING TRADITIONAL AND INTELLIGENT APPROACHES

 

Nitin Raval, Dhaval Parikh

 

This paper investigates the feasibility and viability of adaptive recommender systems by comparing traditional recommendation approaches with a proposed adaptive hybrid model that integrates Graph Neural Networks (GNN), Large Language Models (LLM), and Explainable AI (XAI). The study explores how these components contribute to handling cold-start problems, real-time user preference shifts, and enhancing explainability in recommendation systems. Benchmarking against traditional methods and popular platforms (e.g., Netflix, Spotify), this paper presents a comprehensive comparative analysis and discusses the future of adaptive RecSys for academia and industry. Real-world scenarios are considered, based on simulated benchmarks such as genre-drifted MovieLens, personalized music preferences, and multilingual regional content, demonstrating the practical value and implementation roadmap.

 

Cite: Nitin Raval, Dhaval Parikh FEASIBILITY AND VIABILITY OF ADAPTIVE RECOMMENDER SYSTEMS: BRIDGING TRADITIONAL AND INTELLIGENT APPROACHES. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 694-702, DOI: https://doi.org/10.24412/1932-2321-2025-489-694-702

 

RELIABILITY AND RISK MODELLING IN ECONOMIC, ENVIRONMENTAL SUSTAINABILITY, AND FINANCIAL SYSTEMS WITH THE ARCTAN MARSHALL-OLKIN WEIBULL DISTRIBUTION

 

Moin Uddin

 

The Arctan Marshall-Olkin family is presented as a novel and versatile class of heavy-tailed distributions for modelling extreme occurrences in the fields of economics, finance, and the Environmental sustainability. This family is produced by combining the Arctan-X method, which makes use of the arctangent inverse trigonometric function, with the Marshall-Olkin framework. A specific example, the Arctan-Marshall-Olkin-Weibull (ATMOW) distribution, is thoroughly examined. By adding a third parameter, ATMOW improves its capacity to handle heavy-tailed data in contrast to the traditional two-parameter Weibull distribution. Additionally, to evaluate severe financial risks, closed-form formulas for important actuarial risk measures such as value at risk and tail value at risk are generated. ATMOW performs better than a number of rival multiparameter distributions, according to empirical applications to actual financial, environmental sustainability, and economic datasets.

 

Cite: Moin Uddin RELIABILITY AND RISK MODELLING IN ECONOMIC, ENVIRONMENTAL SUSTAINABILITY, AND FINANCIAL SYSTEMS WITH THE ARCTAN MARSHALL-OLKIN WEIBULL DISTRIBUTION. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 703-715, DOI: https://doi.org/10.24412/1932-2321-2025-489-703-715

 

AUTHENTICATION APPROACHES IN IOT: A COMPARATIVE STUDY OF EXISTING METHODS AND FUTURE DIRECTIONS

 

Dr. Mihir Mehta, Dr.Sanjay Patel, Nrupesh Shah

 

The Internet of Things (IoT) marks a new era in how humans interact with technology. It enables physical devices to generate, receive, and effortlessly exchange data with each other. The primary objective of IoT-based systems is to enhance user convenience and operational efficiency. However, IoT environments are typically open and interconnected, which exposes them to a wide range of security threats. Ensuring robust security is therefore a critical aspect of IoT networks. Traditional security mechanisms are often unsuitable due to the limited computational resources of IoT devices. Authentication plays a key role in verifying the identity of each device within the network, as compromised devices can significantly disrupt operations. To fully leverage IoT's potential, it is crucial to address and resolve security concerns across all layers. The research is divided into two primary categories: authentication factors and widely used cryptographic primitives. Authentication factors encompass methods like passwords, RFID, smart cards, and one-time passwords (OTPs), whereas cryptographic primitives involve techniques such as Data Encryption Standards (DES), Advance Encryption Standard (AES) and Physically Unclonable Functions (PUFs). This paper reviews existing authentication techniques and provides the comparison among them. Our research works also highlights the performance of various cryptographic techniques in IoT network. Reliability denotes the consistent and accurate functioning of an authentication system in verifying that only authorized devices and users are granted access to the IoT network or its associated services.Reliable Authentication framework offers consistent Identity verification & minimizes False Positive- Unauthorized devices mistakenly authenticated/ False Negatives-Legitimate devices wrongly denied access. It concludes with identification of various open research challenges associated with IoT Authentication.

 

Cite: Mihir Mehta, Sanjay Patel, Nrupesh Shah AUTHENTICATION APPROACHES IN IOT: A COMPARATIVE STUDY OF EXISTING METHODS AND FUTURE DIRECTIONS. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 716-730, DOI: https://doi.org/10.24412/1932-2321-2025-489-716-730

 

THE ALDAM-LOG-LOGISTIC DISTRIBUTION AND ITS APPLICATION TO BIMODAL DATA

 

Idzhar A. Lakibul

 

The log-logistic distribution is a commonly used lifetime probability distribution, particularly for modeling lifetime data. In this paper, a new generated (G) family of distributions is introduced, referred to as the Aldam-G family of distributions. This family is derived using the T-X family of distributions. As an application, the log-logistic distribution is generalized through the proposed family, resulting in a new distribution called the Aldam-Log-logistic distribution. The probability density function (PDF) and cumulative distribution function (CDF) of the proposed distribution are derived. It is observed that the new distribution can effectively model bimodal data. Several statistical properties of the Aldam-Log-logistic distribution are also explored. Parameter estimation is carried out using the maximum likelihood method. Finally, the proposed distribution is applied to a real-life dataset, and the results indicate that it provides a better fit compared to some existing competing distributions.

 

Cite: Idzhar A. Lakibul THE ALDAM-LOG-LOGISTIC DISTRIBUTION AND ITS APPLICATION TO BIMODAL DATA. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 731-743, DOI: https://doi.org/10.24412/1932-2321-2025-489-731-743

 

ALPHA LOGARITHMIC TRANSFORMATION OF LOMAX DISTRIBUTION WITH PROPERTIES AND APPLICATION TO SURVIVAL DATA

 

K.M.Sakthivel, Nandhini V, and Vidhya G

 

Probability distribution plays a vital role in lifetime data analysis, reliability modeling, and survival studies. Classical models often exhibit limitations in flexibility when dealing with diverse real-world datasets.To address this gap, we introduce an extended model that improves the ability to capture various shapes of lifetime distributions.In this paper, we propose a new three-parameter probability distribution, termed the Alpha Logarithmic Transformation of Lomax (ALTLx) distribution, which proves to be effective for modeling lifetime data.The proposed distribution is studied in detail, and several of its statistical properties are derived, including moments, incomplete moments, quantiles, mean residual life, mean inactivity time, entropy, order statistics, and stress strength analysis.Parameter estimation is carried out using the method of maximum likelihood to ensure reliable inference. The practical applicability of the ALTLx distribution is demonstrated by fitting it to datasets representing the survival time of patients. The results indicate that it outperforms several well-known competing distributions, showing that the proposed model provides a significantly better fit. The distribution exhibits greater flexibility and effectively captures diverse patterns in lifetime data. The findings confirm that the ALTLx distribution offers superior performance in modeling lifetime and reliability data.

 

Cite: K.M. Sakthivel, Nandhini V, and Vidhya G ALPHA LOGARITHMIC TRANSFORMATION OF LOMAX DISTRIBUTION WITH PROPERTIES AND APPLICATION TO SURVIVAL DATA. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 744-756, DOI: https://doi.org/10.24412/1932-2321-2025-489-744-756

 

LINEAR PREDICTION OF K-RECORD VALUES FROM GENERALIZED PARETO DISTRIBUTION

 

Laji Muraleedharan and Manoj Chacko

 

In this paper, the upper k-record values arising from a generalized Pareto distribution is considered. After considering the means, variances and covariances of the upper k-record values, the best linear unbiased estimators and best linear invariant estimators of the location and scale parameters of the generalized Pareto distribution is discussed under the assumption that the shape parameter is assumed to be known. The marginal best linear unbiased predictor and best linear invariant predictor of future upper k-record value and the joint best linear unbiased predictor and best linear invariant predictor of a pair of future upper k-record values are also determined. Finally, a real dataset is considered to illustrate the proposed inference procedures developed in this paper.

 

Cite: Laji Muraleedharan and Manoj Chacko LINEAR PREDICTION OF K-RECORD VALUES FROM GENERALIZED PARETO DISTRIBUTION. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 757-771, DOI: https://doi.org/10.24412/1932-2321-2025-489-757-771

 

SOME REFINEMENTS OF INEQUALITIES FOR POLYNOMIALS

 

Aijaz Ahmad, D Venkatesan, Ajiaz Magbool Dar, Aafaq A. Rather, Danish Qayoom, Eid Sadun Alotaibi, Showkat A. Bhat, Mushtaq A. Lone and Jameel Ahmad Ansari

 

The study of inequalities for polynomials plays a central role in mathematical analysis, with numerous results exploring the relationships between a polynomial and its derivative. Over time, several refinements and generalizations have been established, strengthening classical inequalities and extending them to broader settings. In this paper, we present new refinements that further enhance existing results on polynomial inequalities. Our findings not only sharpen earlier theorems, but also provide generalized forms that encompass a wider class of polynomials. The refinements introduced here not only strengthen known inequalities, but also provide a framework that connects different strands of results in this area. This contributes to a deeper understanding of the behavior of polynomials in the complex domain and establishes new directions for further investigation.

 

Cite: Aijaz Ahmad, D Venkatesan, Ajiaz Magbool Dar, Aafaq A. Rather, Danish Qayoom, Eid Sadun Alotaibi, Showkat A. Bhat, Mushtaq A. Lone and Jameel Ahmad Ansari SOME REFINEMENTS OF INEQUALITIES FOR POLYNOMIALS. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 772-780, DOI: https://doi.org/10.24412/1932-2321-2025-489-772-780

 

DECISION MAKING UNDER UNCERTAINTY BASED ON GENERALIZED FERMATEAN FUZZY NUMBERS

 

Tanuj Kumar, Gajraj Singh, Birjesh Kumar

 

It has been observed that Fermatean fuzzy sets, which provide a more comprehensive framework than intuitionistic fuzzy sets, offer significant advantages in decision-making processes. This study first formulates the generalized representation of Fermatean fuzzy numbers, including their membership and non-membership functions as well as the associated arithmetic operations. Furthermore, novel defuzzification and ranking techniques are proposed to convert Fermatean fuzzy information into crisp values. To validate the approach, a statistical application is presented for determining the sample range under generalized trapezoidal Fermatean fuzzy numbers using the proposed defuzzification and ranking methods. Finally, a case study is provided to demonstrate the practical applicability of generalized Fermatean fuzzy numbers in real-world decision-making contexts.

 

Cite: Tanuj Kumar, Gajraj Singh, Birjesh Kumar DECISION MAKING UNDER UNCERTAINTY BASED ON GENERALIZED FERMATEAN FUZZY NUMBERS. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 781-794, DOI: https://doi.org/10.24412/1932-2321-2025-489-781-794

 

PERFORMANCE METRICS OF M/M(A,B)/(2,1)/RV QUEUING SYSTEM WITH RENEGING AND RETENTION OF CUSTOMERS

 

Jenifer Princy P, K Julia Rose Mary

 

In general, the quality of service management ensures the quality of the business outcome. Customer behavior trivially has an impact on it. To obtain a potential output, it is necessary to speed up the customer service request. Adequate among of losses happen when the customers withdraw the service request. Thus, it is vital to implement a significant strategy based on the number of request in queue. By modeling the service system mathematically in concern with the customers behavior, it is elegant to make a strategy for estimating demands. This paper aims to interpret a system, M/M(a,b)/(2,1)/RV providing service in batches with the aid of two servers and encounters the measures of performance of the system under the possibility of servers vacation and customers behavior of reneging and retention. A server

withdraws the system whenever he is not available with an accessible batch to start his service. This behavior of the server can be termed as vacation. The vacation policy used in this paper involves the repeated mode of vacation of the servers. Customers entering the system experience three states of availability of the servers. Thus, for the proposed model the equations and solutions for the steady state is derived and computed theoretically, and the results are verified through numerical illustration.

 

Cite: Jenifer Princy P, K Julia Rose Mary PERFORMANCE METRICS OF M/M(A,B)/(2,1)/RV QUEUING SYSTEM WITH RENEGING AND RETENTION OF CUSTOMERS. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 795-805, DOI: https://doi.org/10.24412/1932-2321-2025-489-795-805

 

FM/FM/1 QUEUE WITH N-POLICY TWO-PHASE, SERVER START-UP, TIME-OUT AND BREAKDOWNS USING L-R METHOD

 

John Benhur.K, D.Gopinath, P.Sudhakara Babu, Muragesh Math

 

This study centers on evaluating the performance metrics of the system, specifically emphasizing two key elements:the expected length of the system (L) and the minimum anticipated cost T(N) derived from the optimal strategy analysis of the N-policy L-R(Left-Right).FM/FM/I two-phase queueing model.This system incorporates essential factors such as server start-up,time-out,breakdown, and repair processes,making the analysis more comprehensive and reflective of real-world operational challenges.The arrival rate and service rate in this model are represented as triangular fuzzy numbers,which provide a more flexible and realistic approach in capturing the uncertainty and variability inherent in these rates.The study employs the L-R method,a technique known for its efficiency and simplicity when compared to the more traditional alpha-cuts method.The L-R method minimizes computational complexity and enhances the case of differentiation, making it a more convenient approach for handling fuzzy numbers in queueing theory.By applying this method,the study effectively addresses the limitations of previous models and provides a more accurate estimation of the system's performance metrics.To demonstrate the validity and applicability of the proposed model,a numerical example is provided.This example illustrates the practicality of the L-R method in optimizing the system's cost and performance,shoocasing how the model can be applied to real-world scenarios.The results of the numerical example confirm the robustness and reliability of the proposed approach,making a strong case for its adoption in queueing system analysis.This study thus offers significant insights and a novel methodological contribution to the field of fuzzy queueing systems.

 

Cite: John Benhur.K, D. Gopinath, P.Sudhakara Babu, Muragesh Math FM/FM/1 QUEUE WITH N-POLICY TWO-PHASE, SERVER START-UP, TIME-OUT AND BREAKDOWNS USING L-R METHOD. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 806-813, DOI: https://doi.org/10.24412/1932-2321-2025-489-806-813

 

A COMPREHENSIVE EXPLORATION OF QUASI RAMA DISTRIBUTION WITH SIGNIFICANT STATISTICAL PROPERTIES AND APPLICATIONS

 

Rashid A. Ganaie, Ashok Routhu, T. Vivekanandan, Manzoor A. Khanday and Prem Lata Gautam

 

In this study, we propose a novel extension of the quasi Rama distribution referred to as the length-biased quasi Rama (LBQR) distribution, developed by applying the length-biased technique to the classical quasi Rama distribution. The proposed distribution is thoroughly investigated and several of its important statistical properties are derived including moments, harmonic mean, order statistics, reliability function, hazard rate function, reverse hazard rate function, moment generating function, characteristic function, Renyi entropy as well as Bonferroni and Lorenz curves. Furthermore, the parameters of the new distribution are estimated using the method of maximum likelihood estimation. To evaluate the practical relevance and predictive performance of the length biased quasi Rama distribution, it is applied to two real lifetime data sets, demonstrating its applicability and effectiveness in modeling real-world lifetime phenomena.

 

Cite: Rashid A. Ganaie, Ashok Routhu, T. Vivekanandan, Manzoor A. Khanday and Prem Lata Gautam A COMPREHENSIVE EXPLORATION OF QUASI RAMA DISTRIBUTION WITH SIGNIFICANT STATISTICAL PROPERTIES AND APPLICATIONS. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 814-827, DOI: https://doi.org/10.24412/1932-2321-2025-489-814-827

 

METRIC DIMENSION OF GRAPHS WITH SEQUENTIAL PENDANT EDGE EXTENSION

 

Ajendra Kumar, B.S. Rawat, Kirti Chauhan

 

The metric dimension of a graph measures the smallest number of vertices (landmarks) needed so that every vertex in the graph can be uniquely identified by its distances to these landmarks. In this study, we investigate a specific structural modification called Sequential Pendant Edge Extension (SPEE) in which pendant edges are attached one at a time to an existing graph. We analyse whether such extensions affect the metric dimension of graph families. These results provide both theoretical insights and practical assurance in designing reliable and scalable network models.

 

Cite: Ajendra Kumar, B.S. Rawat, Kirti Chauhan METRIC DIMENSION OF GRAPHS WITH SEQUENTIAL PENDANT EDGE EXTENSION. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 828-838, DOI: https://doi.org/10.24412/1932-2321-2025-489-828-838

 

A REVIEW ON SUSTAINABILITY, RELIABILITY, AND SAFETY OF MICROCHANNEL HEAT SINKS IN ELECTRONIC COOLING APPLICATIONS

 

Santosh Kumar Rai, Pardeep Kumar, Mahesh Kumar Gupta, Dhowmya Bhatt,Arun Uniyal,Gyanendra Prasad Bagri, Nikhil Vivek Shrivas

 

Day by day, the rapid growth in the miniaturization of electronic devices ranging from mini, micro, to nano scales for various applications has drastically increased. However, the fabrication of such devices and the dissipation of heat from their bodies to the surrounding environment remains challenging. Due to these thermal management issues, electronic devices, MEMS, and other microfluidic and biomedical systems face significant sustainability, reliability, and safety concerns. In the present paper, we address various methods to overcome these problems in existing devices by using microchannels (MICHs). MICHs are used to remove heat from electronic components, thereby improving device safety and enabling them to operate within permissible temperature limits. A comprehensive literature review is presented, covering the classification of MICHs, their fabrication methods, and their applications in enhancing the reliability and safety of various devices. Sustainable design aspects are also considered to improve energy efficiency, optimize material usage, reduce the environmental impact of coolants, and enhance product life cycle performance. Furthermore, the paper outlines potential future research directions and fields of innovation for further advancements in this domain.

 

Cite: Santosh Kumar Rai, Pardeep Kumar, Mahesh Kumar Gupta, Dhowmya Bhatt, Arun Uniyal, Gyanendra Prasad Bagri, Nikhil Vivek Shrivas A REVIEW ON SUSTAINABILITY, RELIABILITY, AND SAFETY OF MICROCHANNEL HEAT SINKS IN ELECTRONIC COOLING APPLICATIONS. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 839-851, DOI: https://doi.org/10.24412/1932-2321-2025-489-839-851

 

INVESTIGATING THE PERFORMANCE OF MAP/PH/1 QUEUEING SYSTEMS WITH CATASTROPHIC FAILURES, NEGATIVE ARRIVALS, SERVER BREAKDOWNS, REPAIRS AND LACK OF CUSTOMER TOLERANCE FOR DELAYS

 

K. Thilagavathy, G. Ayyappan

 

This paper presents a comprehensive analytical study of a single-server queueing system incorporating by a Markovian Arrival Process (MAP), Phase-Type (PH) service times, and multiple complex disruptions, including catastrophic failures, negative arrivals, server breakdowns, and phase-type repairs. Utilizing matrix-analytic methods, we derive steady-state probability distributions, and system performance measures. The consequence of failure frequencies, and repair parameters on key performance metrics such as system size, and server availability is thoroughly examined. The behavior of the system is demonstrated using numerical results. The results offer valuable insights for the purposes of design and optimization of resilient and efficient service systems operating in uncertain and failure-prone environments.

 

Cite: K. Thilagavathy, G. Ayyappan INVESTIGATING THE PERFORMANCE OF MAP/PH/1 QUEUEING SYSTEMS WITH CATASTROPHIC FAILURES, NEGATIVE ARRIVALS, SERVER BREAKDOWNS, REPAIRS AND LACK OF CUSTOMER TOLERANCE FOR DELAYS. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 852-866, DOI: https://doi.org/10.24412/1932-2321-2025-489-852-866

 

BAYESIAN STUDY OF HYBRID WEIBULL- EXPONENTIAL POWER MODEL BY MEANS OF HAMILTONIAN MONTE CARLO ALGORITHM

 

M. A. Alhassan, A. Yahaya, O. O. Ishaq, B. Abba, A. B. Muhammad

 

In this paper, we derived two important properties, viz: moment and order statistics, for the Hybrid Weibull-Exponential Power actuarial reliability model. We considered Bayesian procedures for estimating the parameters of the model. We tested the performance of the estimation approach by means of simple Monte Carlo experiment, hence, found the estimator consistent for the model parameters. Subsequently, we utilized the Bayesian inferential technique to further explore the overall adequacy the model in real—life through two sets of failure time reliability data. However, at the complexity point of the Bayesian method in dealing with the complex posterior of the model, we ushered-in the robust Hamiltonian Monte Carlo simulation algorithm, and thus, swiftly enhanced the posterior inference. The Add-on MLE examination on the model parameters was briefly conducted to additionally check the model's performance in real life. On the premise of the evaluation criteria and goodness-of-fit statistics results obtained from the MLE computation, we, once again, found clearly that the model is superior among the similar models compared. The HWEP model is, thus, apt for reliability analyses of dual failures as well as dual-component failed systems.

 

Cite: M. A. Alhassan, A. Yahaya, O. O. Ishaq, B. Abba, A. B. Muhammad BAYESIAN STUDY OF HYBRID WEIBULL- EXPONENTIAL POWER MODEL BY MEANS OF HAMILTONIAN MONTE CARLO ALGORITHM. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 867-880, DOI: https://doi.org/10.24412/1932-2321-2025-489-867-880

 

PERFORMANCE AND COST ANALYSIS OF A GI/M(a,b)/I QUEUE WITH COMPULSORY AND EXTENDED REPAIRS DURING MULTIPLE WORKING VACATIONS

 

E. Praveena, K. Julia Rose Mary

 

The paper analyses a single-server queuing model with general distribution arrival and batch service under multiple working vacations, incorporating both compulsory and extended repairs. Customers receive service in groups following a general bulk service approach, instead of being served one at a time. Each service batch requires a minimum of ‘a’ and maximum of ‘b’ customers. The study considers two types of repairs—compulsory and extended—and derives the steady-state equations, performance measures, and expected cost of the model. Finally, the impact of various parameters is illustrated in numerical analysis section and graphical representation.

 

Cite: E. Praveena, K. Julia Rose Mary PERFORMANCE AND COST ANALYSIS OF A GI/M(a,b)/I QUEUE WITH COMPULSORY AND EXTENDED REPAIRS DURING MULTIPLE WORKING VACATIONS. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 881-894, DOI: https://doi.org/10.24412/1932-2321-2025-489-881-894

 

SURVIVAL ANALYSIS OF AREA BIASED GENERALIZED SAULEH DISTRIBUTION WITH STATISTICAL PROPERTIES AND APPLICATIONS

 

Dr. V. Arulmozhi, Dr. R. Sivaranjani, Dr. K. Poovizhi

 

The "Sauleh distribution" is a novel lifespan distribution that is suggested by this study. Among the statistical and mathematical features that have been discussed are the PDF and CDF. Numerous helpful statistical features are derived, including the Area Biased Generalized Sauleh distribution's hazard rate function, reverse hazard rate function, order statistics, moments, measure of skewness, and measure of kurtosis. It uses the well-known maximum likelihood outcome approach to estimate the parameter of the distribution. To demonstrate this distribution's adaptability and superiority over alternative distributions, a comparative analysis is conducted in the end. It establishes its significance in reliability engineering by producing a rich class of distributions that may capture various shapes and behaviors and, as a result, enable better fitting to empirical data. Real-world data sets are used to establish the use of this novel distribution. Melanoma patient survival data has demonstrated the effectiveness of the proposed distribution; it is concluded that the produced distribution provides a superior fit.

 

Cite: Dr. V. Arulmozhi, Dr. R. Sivaranjani, Dr. K. Poovizhi SURVIVAL ANALYSIS OF AREA BIASED GENERALIZED SAULEH DISTRIBUTION WITH STATISTICAL PROPERTIES AND APPLICATIONS. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 895-904, DOI: https://doi.org/10.24412/1932-2321-2025-489-895-904

  

OPTIMIZATION OF A FUZZY SUPPLY CHAIN INVENTORY SYSTEM FOR DETERIORATING ITEMS UNDER ENVIRONMENTAL SUSTAINABILITY

 

Vikash, Dhirendra Singh, Ashish Negi

 

The growing emphasis on environmental sustainability and operational reliability has encouraged researchers to extend classical inventory models using fuzzy logic to better handle uncertainty. This study proposes a sustainable inventory framework for deteriorating items, where demand is influenced by both selling price and greening initiatives. The model incorporates investments in preservation technology and green practices under trapezoidal fuzzy environments. Shortages are permitted, and the problem is solved through signed distance defuzzification method, supported by MATLAB for computation and analysis. To demonstrate the applicability, numerical examples are presented, followed by a comparative analysis between crisp and fuzzy settings. The results show that the fuzzy-based model consistently reduces overall costs, with an average decrease of 4.72%, while also lowering cycle time and reorder-to-shortage intervals by 4.04% and 5.52%, respectively. Sensitivity analysis further reveals that total cost is most sensitive to fluctuations in green technology and procurement expenses, whereas preservation and shortage costs have comparatively smaller effects. The findings highlight that adopting fuzzy-based decision approaches not only enhances financial performance but also supports ecological objectives, offering a more reliable pathway for sustainable supply chain management.

 

Cite: Vikash, Dhirendra Singh, Ashish Negi OPTIMIZATION OF A FUZZY SUPPLY CHAIN INVENTORY SYSTEM FOR DETERIORATING ITEMS UNDER ENVIRONMENTAL SUSTAINABILITY. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 905-920, DOI: https://doi.org/10.24412/1932-2321-2025-489-905-920

 

STATISTICAL INFERENCE ON THE GENERALIZED EXPONENTIAL DISTRIBUTION BASED ON GENERALIZED ORDER STATISTICS CHARACTERIZATIONS AND APPLICATIONS

 

Imtiyaz A. Shah and Khalid Ul Islam Rather

 

In this paper, we study the Some distribution functions have been characterized based on m-dual generalized order statistics and consequently m-generalized order statistics. Moreover, we show that these characterization properties provide a beneficial strategy to predict future events, which are based on past or current events and on an arbitrary distribution function. Finally, an application of these results is given for bivariate generalized exponential distribution.

 

Cite: Imtiyaz A. Shah and Khalid Ul Islam Rather STATISTICAL INFERENCE ON THE GENERALIZED EXPONENTIAL DISTRIBUTION BASED ON GENERALIZED ORDER STATISTICS CHARACTERIZATIONS AND APPLICATIONS. DOI: https://doi.org/10.24412/1932-2321-2025-489-921-930

 

PARAMETRIC ESTIMATION OF THE PROCESS CAPABILITY INDEX S''pk AND ITS APPLICATION TO ELECTRONIC INDUSTRIES

 

Mahendra Saha, Pratibha Pareek, Anju Devi, Abhimanyu S Yadav

 

The proposed index is the process capability index used in the electronics industries to measure the capability of the process. This article focuses on process capability index, specifically applicable to normal random variables. The article has three main objectives: Firstly, we explore various classical estimation methods for the proposed index from frequentist approaches for normal distributions and compare their performance based on mean squared errors. Second, we calculate the classical confidence interval for the proposed index, which includes the asymptotic confidence interval. Third, we examine both Bayes point

and interval estimation under symmetric and asymmetric loss functions for the proposed index. A Monte Carlo and Markov Chain Monte Carlo simulation study is conducted to compare the performance of the classical and the Bayes estimates of the proposed index for some set of parameters. Finally, to demonstrate the applicability of this index, two real data sets from the electronics industry are re-analyzed.

 

Cite: Mahendra Saha, Pratibha Pareek, Anju Devi, Abhimanyu S Yadav PARAMETRIC ESTIMATION OF THE PROCESS CAPABILITY INDEX S''pk AND ITS APPLICATION TO ELECTRONIC INDUSTRIES. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 931-947, DOI: https://doi.org/10.24412/1932-2321-2025-489-931-947

 

TWO COMPONENT MIXTURE OF LEFT TRUNCATED THREE PARAMETER LOGISTIC TYPE DISTRIBUTION WITH APPLICATIONS TO MANPOWER MODELING

 

Usha Hamsa, Srinivasa Rao Kraleti, Muniswamy Begari

 

The logistic distribution serves as an alternative to normal distribution due to its is similarity in shape with heavier tails. Several extensions and generalizations of the logistic distribution have been proposed and explored in research papers published from 2020 onwards, particularly for modeling data with varying characteristics like skewness and kurtosis, or for specific applications in areas like finance and survival analysis. Some important distributions include the skew logistic distribution, the extended log-logistic distribution, two parameter logistic type distribution, three parameter logistic type distribution and the beta log-logistic distribution. These extensions offer increased

flexibility in fitting data compared to the standard logistic distribution. Recently Usha et al (2025) have studied truncated three parameter Logistic distribution. Based on this distribution, we propose a new model called the two component mixture of left truncated tree parameter logistics type distribution for describing manpower data. It involves various specialized models, such as mixture of truncated logistic and a mixture of two-parameter logistic type distributions, among several others, so it's possibly better model for analyzing skewed data. The new density function can be expressed as a weighted combination of truncated tree parameter logistic-type densities. Various mathematical properties of the new distribution including explicit expressions for the moments are derived. We discuss the maximum likelihood method to estimate the model parameters for heterogeneous data. The superiority of the proposed manpower model is illustrated.

 

Cite: Usha Hamsa, Srinivasa Rao Kraleti, Muniswamy Begari TWO COMPONENT MIXTURE OF LEFT TRUNCATED THREE PARAMETER LOGISTIC TYPE DISTRIBUTION WITH APPLICATIONS TO MANPOWER MODELING. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 948-968, DOI: https://doi.org/10.24412/1932-2321-2025-489-948-968

 

RISK ANALYSIS OF A REPAIRABLE UNIT IN SUGAR INDUSTRY: A HYBRID APPROACH USING FUZZY FMEA AND GRA

 

Anushree and Seema Sharma

 

The purpose of this study is to develop a comprehensive risk assessment profile for a repairable unit in a sugar industry. The risk analysis of milling unit in the sugar industry often relies on uncertain data and subjective expert assessments, which could be biased. This highlights the need for careful interpretation as the data quality variations have great impact on the decision-making. By integrating risk analysis techniques like failure mode and effect analysis (FMEA), Grey relation analysis (GRA), and fuzzy theory, together with expert assessments for the very first time, the approach aims to prioritize risk management strategies of milling unit in Dhampur Sugar Mills Ltd., a sugar industry located in Western Ultrar Pradesh, India. First of all, Root cause analysis (RCA) is conducted to detect in-depth potential failure modes of the unit. Further, FMEA is employed for risk identification and prioritization, with ranking of components based on their RPN (Risk Priority Number) scores. To address

the limitations of FMEA, fuzzy FMEA supported by a fuzzy decision support system (FDSS) has been used to compute FRPN (Fuzzy Risk Priority Number) scores. Based on these scores, the components are ranked. Subsequently, GRA is employed within the fuzzy FMEA framework to intelligently assess the relative importance of key factors, i.e. probability of occurrence, severity effect, and probability of non-detectability of failure in order to determine optimal rankings of critical components. The implementation of fuzzy logic and grey approach with inputs of FMEA offers a more efficient and better way to make decision regarding risk priorities. This risk assessment approach is valuable for sugar industry where complex engineering systems are essential for production process, and the findings from this study will aid system specialists and analysis in future planning and maintenance strategies.

 

Cite: Anushree and Seema Sharma RISK ANALYSIS OF A REPAIRABLE UNIT IN SUGAR INDUSTRY: A HYBRID APPROACH USING FUZZY FMEA AND GRA. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 969-988, DOI: https://doi.org/10.24412/1932-2321-2025-489-969-988

 

BAYESIAN SINGLE SAMPLING PLANS WITH ZERO-INFLATED BINOMIAL DISTRIBUTION USING COST OPTIMIZATION

 

Sangeetha S, Shalini K and Hemalatha R

 

The economic design of sampling plans involves determining an optimal plan that minimizes the total cost associated with inspection while ensuring acceptable product quality levels. Bayesian sampling plans are better suited for handling significant between-lot variations in quality, leading to more informed acceptance decisions due to their capability to incorporate prior information. The objective of this article is to design Bayesian single sampling plans considering a Beta prior and a Zero-inflated Binomial (ZIB) distribution, using cost optimization principles by developing an economic model aimed at achieving optimal total costs, considering the Average Total Inspection (ATI). A numerical illustration is provided to demonstrate the selection of Bayesian single sampling plans under the ZIB distribution that minimizes the producer’s total cost.

 

Cite: Sangeetha S, Shalini K and Hemalatha R BAYESIAN SINGLE SAMPLING PLANS WITH ZERO-INFLATED BINOMIAL DISTRIBUTION USING COST OPTIMIZATION. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 989-995, DOI: https://doi.org/10.24412/1932-2321-2025-489-989-995

 

MODELING STRENGTH OF THE AIRCRAFT WINDOW GLASS BY TRUNCATED SHAMBHU DISTRIBUTION

 

Dhanashri Patil, Manjushri Chikkalli, Shruti Hipale, Chandrakant Gardi

 

Failure of regular distributions to appropriately model the real life scenarios demands the new distributions, which will be more useful. Truncated distributions are found to model many real datasets more precisely than corresponding regular distribution. In this paper, a novel truncated Shambhu distribution (TS) is proposed to model the data, which is naturally constrained by physical, financial, or environmental factors. Some statistical properties including moments, survival and hazard functions, are discussed. For better understanding of the proposed distribution, plots of the distribution are provided. Estimation of parameters is discussed using the maximum likelihood estimation method. Finally, an application based on real data is considered to illustrate the practical relevance of the proposed distribution.

 

Cite: Dhanashri Patil, Manjushri Chikkalli, Shruti Hipale, Chandrakant Gardi MODELING STRENGTH OF THE AIRCRAFT WINDOW GLASS BY TRUNCATED SHAMBHU DISTRIBUTION. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 996-1005, DOI: https://doi.org/10.24412/1932-2321-2025-489-996-1005

 

RELIABLE SINGLE DIODE PATCH ANTENNA WITH DUAL MODE FREQUENCY AND PATTERN RECONFIGURABILITY WITH COMPLIANCE ANALYSIS

 

Sameer Mansuri, Shahid S. Modasiya

 

This study introduces a compact rectangular patch antenna that achieves frequency reconfiguration using a single PIN diode, enabling two distinct operational modes. The design integrates multiple strategically placed slots within the patch to support a wide range of wireless communication bands. By employing only one diode, the antenna reduces circuit complexity and fabrication cost while ensuring efficient performance. The optimization was carried out through full-wave electromagnetic simulations, with reflection coefficients and radiation patterns serving as primary evaluation metrics. A prototype fabricated on a cost-effective FR4 substrate was tested in an anechoic chamber, showing strong correlation between simulated and measured results, with effective impedance matching. Furthermore, reliability assessments and electromagnetic safety evaluations confirmed compliance with international SAR standards and consistent performance under environmental stress conditions. This work highlights that carefully designed slot integration on FR4 substrates can provide performance levels comparable to those of specialized high-frequency materials, delivering a compact and economical solution for modern communication systems.

 

Cite: Sameer Mansuri, Shahid S. Modasiya RELIABLE SINGLE DIODE PATCH ANTENNA WITH DUAL MODE FREQUENCY AND PATTERN RECONFIGURABILITY WITH COMPLIANCE ANALYSIS. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 1006-1016, DOI: https://doi.org/10.24412/1932-2321-2025-489-1006-1016

 

STATISTICAL MODELLING AND APPLICATIONS OF THE ET–EXPONENTIAL DISTRIBUTION: EVIDENCE FROM BIOMEDICAL AND ENGINEERING DATA

 

T. Kamaleshwar, R. Seethalakshmi and S. Mythily

 

The current study utilizes the exponential distribution as the baseline cumulative function and applies the T-x transformation on the exponential family in order to develop a new class of lifetime distribution. The main objective of this study is to develop and evaluate a flexible distribution that may capture various kinds of lifespan features present in real-world applications. A specific instance, the ET-Exponential (ET-Exp) distribution, is thoroughly examined. The structural properties of the paper, such as the moments, cumulative distribution function, and probability density function, are derived. Parameter estimation is performed via the maximum likelihood method. The suggested approach is used on real datasets, such as carbon fiber breaking stress data and hypertension survival times. The performance of the model is assessed by using statistical criteria such as Akaike Information criteria (AIC), Bayesian Information Criteria (BIC), Kolmogorov-Simron (KS), and Anderson-Darling (AD) tests, which were among the statistical tests used to estimate the goodness of fit of each distribution's essential parameters. The results demonstrate that the proposed ET-Exponential distribution is a flexible and effective tool for modeling lifetime data and shows promise for applications in both biomedical and industrial fields.

 

Cite: T. Kamaleshwar, R. Seethalakshmi and S. Mythily STATISTICAL MODELLING AND APPLICATIONS OF THE ET–EXPONENTIAL DISTRIBUTION: EVIDENCE FROM BIOMEDICAL AND ENGINEERING DATA. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 1017-1026, DOI: https://doi.org/10.24412/1932-2321-2025-489-1017-1026

 

EIGHT YEARS OF RESEARCH IN RELIABILITY: THEORY AND APPLICATIONS (2017–2024): A SCOPUS BASED BIBLIOMETRIC ANALYSIS

 

Sajad Hussain, Vilayat Ali Bhat

 

This study looks at the SCOPUS research published in the journal Reliability. Theory and Applications from 2017 to 2024. The main goal is to understand how the journal has grown, how influential it is, and how researchers are working together. The data shows that the number of papers published has increased steadily, especially after 2020, showing that the journal is becoming more important in the academic world. Most of the research came from countries like India, the Russian Federation, and Azerbaijan. Universities such as Kadyrov Chechen State University and Azerbaijan State Oil and Industry University were among the top contributors. The study also used network analysis to examine how authors collaborate, which keywords are used together, and which papers are cited the most. It was found that a few researchers, like Yusuf Ibrahim and Rather Aafaq A., are very active and well-connected in the author network. Citation analysis showed which authors, institutions, and countries had the most influence. This study provides helpful information for researchers, universities, and funding agencies interested in working with or contributing to this journal.

 

Cite: Sajad Hussain, Vilayat Ali Bhat EIGHT YEARS OF RESEARCH IN RELIABILITY: THEORY AND APPLICATIONS (2017–2024): A SCOPUS BASED BIBLIOMETRIC ANALYSIS. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 1027-1045, DOI: https://doi.org/10.24412/1932-2321-2025-489-1027-1045

 

STRESS-STRENGTH MODELLING: A COMPARATIVE STUDY USING SRS, RSS AND MRSS FOR INVERSE NAKAGAMI DISTRIBUTION

 

Surinder Kumar, Rahul Shukla, Bhupendra Meena,Shivendra Pratap Singh

 

The present research endeavours to use the ranked set sampling technique in place of the simple random sampling technique in estimating the reliability model in the field of reliability engineering and manufacturing industries. The stress-strength reliability is widely used in all field of engineering and in the manufacturing industry. In this research article, we take the case when the lifetime data of any machinery followed the Inverse Nakagami distribution, and by considering the stress-strength model, we compare the standard ranked set sampling and median ranked set sampling estimate with the corresponding simple random sampling estimate through Monte Carlo simulation technique. The final stage of this study involves corroborating the findings through an analysis of a benchmark dataset on elevators and their motors utilized in farming machinery.

 

Cite: Surinder Kumar, Rahul Shukla, Bhupendra Meena, Shivendra Pratap Singh STRESS-STRENGTH MODELLING: A COMPARATIVE STUDY USING SRS, RSS AND MRSS FOR INVERSE NAKAGAMI DISTRIBUTION. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 1046-1067, DOI: https://doi.org/10.24412/1932-2321-2025-489-1046-1067

 

DOUBLE ORBIT RETRIAL QUEUEING-INVENTORY SYSTEM WITH ORBITAL SEARCH UNDER (s, S) POLICY

 

M. Thilakavathy, Agassi Melikov, G. Ayyappan

 

We consider two-orbit retrial queueing inventory system with a single server. Service time follows phase-type distribution, while arrivals follow the Markovian arrival process. During busy period of the server, the arriving customer may either balk or move to the limitless orbit. Following service completion, the server takes a vacation if there are no customers in the system. When a breakdown happens during busy times, the server undergoes phase-type repair, and the current customer enters finite orbit. Once the repair is finished, the server either stays idle or searches the finite orbit. The replenishment period is exponential and we use (s, S) policy. The matrix analytic approach has been used to analyze steady state probability. We assess this retrial queueing inventory model both numerically and graphically.

 

Cite: M. Thilakavathy, Agassi Melikov, G. Ayyappan DOUBLE ORBIT RETRIAL QUEUEING-INVENTORY SYSTEM WITH ORBITAL SEARCH UNDER (s, S) POLICY. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 1068-1082, DOI: https://doi.org/10.24412/1932-2321-2025-489-1068-1082

 

RISK MODELING WITH THE NEW LOGARITHMIC TANGENT WEIBULL DISTRIBUTION: APPLICATIONS IN INSURANCE AND NET OFFICIAL DEVELOPMENT ASSISTANCE

 

Palanisamy Manigandan, D. Kanagajothi, R. Jeena and Gulasal Madrakhimova

 

The Weibull distribution is widely used for modeling lifetime and reliability data but often struggles with complex hazard structures and heavy-tailed behaviors. To address these limitations, this study proposes the New Logarithmic Tangent Weibull (NLT-Wei) distribution, which integrates logarithmic and tangent transformations into the Weibull model, introducing an additional shape parameter to enhance flexibility. Key functions, including the PDF, CDF, survival, and hazard functions, are derived, along with quantiles and reliability measures. Parameters are estimated using maximum likelihood estimation (MLE). The model is applied to two real datasets: insurance survival data and Net Official Development Assistance (Net ODA) data. Comparative analyses with Weibull, Exponentiated Weibull, Flexible Weibull, NGLog-Weibull, and Kumaraswamy Weibull distributions are conducted using information criteria (AIC, BIC, CAIC, HQAIC) and goodness-of-fit tests (KS, Cramer-von Mises, Anderson-Darling). Results show that the NLT-Wei consistently achieves lower information criteria, smaller test statistics, and higher p-values, confirming superior fit. Its flexibility allows effective modeling of monotone, bathtub, and inverted-bathtub hazard rates, as well as heavy-tailed behaviors. The study demonstrates that the NLT-Wei is not only robust for reliability and survival analysis but also highly suitable for actuarial and financial applications, offering a practical tool for risk assessment, insurance modeling, and lifetime data analysis across diverse domains.

 

Cite: Palanisamy Manigandan, D. Kanagajothi, R. Jeena and Gulasal Madrakhimova RISK MODELING WITH THE NEW LOGARITHMIC TANGENT WEIBULL DISTRIBUTION: APPLICATIONS IN INSURANCE AND NET OFFICIAL DEVELOPMENT ASSISTANCE. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 1083-1093, DOI: https://doi.org/10.24412/1932-2321-2025-489-1083-1093

 

STATISTICAL APPLICATIONS OF THE KUMARASWAMY ALPHA-POWER LOG-LOGISTIC DISTRIBUTION: EVIDENCE FROM HYPERTENSION AND CO2 EMISSIONS IN INDIA

 

A. Ashok, D. Kanagajothi, T. Kamaleshwar, Palanisamy Manigandan, Abdullayeva Shakhnoza Yerkinovna

 

This article introduces a novel statistical model, the Kumaraswamy Alpha-Power Log-Logistic (KAPLL) distribution, as a flexible extension of the traditional log-logistic model. The KAPLL distribution is especially well-suited for examining a variety of ageing and failure events as it is made to improve modelling capabilities by supporting both symmetric and asymmetric forms. Reliability analysis and lifespan data modelling are made more flexible by the KAPLL model's capacity to represent its probability density function as a combination of log-logistic densities. The hazard rate function (HRF), survival function, moments, quantile function (QF), and moment-generating function (MGF) are among the statistical and mathematical characteristics of the KAPLL distribution that are comprehensively examined. The estimation of KAPLL parameters is achieved through maximum likelihood estimation (MLE), a widely used statistical method. We use the KAPLL distribution to measures of hypertension and carbon dioxide (CO₂) emissions in order to demonstrate the usefulness of the suggested model. The findings demonstrate the KAPLL distribution's adaptability and better fit when compared to other well-known log-logistic model extensions, confirming its promise as a useful tool for modelling complicated behaviours in environmental and biological data.

 

Cite: A. Ashok, D. Kanagajothi, T. Kamaleshwar, Palanisamy Manigandan, Abdullayeva Shakhnoza Yerkinovna STATISTICAL APPLICATIONS OF THE KUMARASWAMY ALPHA-POWER LOG-LOGISTIC DISTRIBUTION: EVIDENCE FROM HYPERTENSION AND CO2 EMISSIONS IN INDIA. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 1094-1103, DOI: https://doi.org/10.24412/1932-2321-2025-489-1094-1103

 

VARIABLES CHAIN SAMPLING PLAN BASED ON THE MINIMUM ANGLE METHOD INVOLVING MINIMUM SUM OF RISKS

 

P. Suthersan and S. Balamurali

 

This study presents a designing methodology for a variables based chain sampling plan utilizing the minimum angle criterion. In the context of acceptance sampling, managing both producer’s risk and consumer’s risk is essential for ensuring consistent product quality, particularly within industrial applications. These risks must be addressed concurrently when making quality related decisions. In this work, both risks are effectively reduced by minimizing the tangent angle formed between two points on the operating characteristic curve: the producer’s quality level, associated with the probability of accepting a good lot, and the consumer’s quality level, associated with the probability of accepting a bad lot. The variables chain sampling plan provides the underlying framework for this formulation. The proposed design procedures are detailed through step-by-step methodologies to demonstrate the implementation of the approach.

 

Cite: P. Suthersan and S. Balamurali VARIABLES CHAIN SAMPLING PLAN BASED ON THE MINIMUM ANGLE METHOD INVOLVING MINIMUM SUM OF RISKS. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 1104-1114, DOI: https://doi.org/10.24412/1932-2321-2025-489-1104-1114

 

MARKOV CHAIN MONTE CARLO AND MODERN BAYESIAN COMPUTATION

 

Manzoor A. Khanday, Shaik Afsar Jahan, Isha Gilhotra, Dilawar A. Bhat

 

Bayesian inference offers a robust approach to the inclusion of prior knowledge in statistical modeling and uncertainty quantification in parameter estimates. Exact Bayesian computation is usually intractable for sophisticated models, and approximation methods are used instead. Markov Chain Monte Carlo (MCMC) techniques are now the foundation of contemporary Bayesian computation and facilitate efficient sampling from high-dimensional posterior distributions. This paper discusses main MCMC algorithms, such as Metropolis-Hastings and Gibbs Sampling, and latest developments like the No-U-Turn Sampler (NUTS). By discussing practical use cases, logistic regression, time series prediction, hierarchical modeling, and count regression data, we illustrate how MCMC-based methods can be used to tackle real-world problems. Each use case takes advantage of PyMC’s MCMC engine to get posterior samples and model performance. Convergence diagnostics, posterior uncertainty, and model interpretability are covered at length. The paper is both a theoretical and practical treatise on MCMC in the context of contemporary Bayesian computation.

 

Cite: Manzoor A. Khanday, Shaik Afsar Jahan, Isha Gilhotra, Dilawar A. Bhat MARKOV CHAIN MONTE CARLO AND MODERN BAYESIAN COMPUTATION. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 1115-1128, DOI: https://doi.org/10.24412/1932-2321-2025-489-1115-1128

 

A REVIEW OF COMPETING AND SEMI-COMPETING RISKS METHODS IN SURVIVAL ANALYSIS

 

Mithali S Kumar, Shruthi P, and Jiju Gillariose

 

Recent advancements in competing and semi-competing risk models have significantly contributed to fields like biomedical, reliability studies (on items related to machines), epidemiological studies (treatment outcomes, disease progression, cause-specific deaths), and so on, with new emerging models. The existing frameworks are also being extended to enhance the theoretical robustness. This article provides a comprehensive review from the very first model to the latest developments, aiming to consolidate the diverse methodologies and offering insights into the most suitable approaches for various practical scenarios. The methodologies adopted in different contexts and the steps that form the base for the study are explained for a better understanding. The methodologies are adopted on the account to the nature of the problem which we are dealing with and in this article the summary of the review is incorporated so that the practitioners can decide which methodology to adopt along with the common pitfalls for a specified case. Domain-specific applications are also incorporated, which helps the practitioners to understand the real-life application of these concepts as well as it paves the way to make further contributions to society in the related fields. By examining both foundational and contemporary contributions, this review seeks to guide researchers and practitioners in selecting effective strategies for handling competing and semi-competing risks in survival analysis.

 

Cite: Mithali S Kumar, Shruthi P, and Jiju Gillariose A REVIEW OF COMPETING AND SEMI-COMPETING RISKS METHODS IN SURVIVAL ANALYSIS. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 1129-1139, DOI: https://doi.org/10.24412/1932-2321-2025-489-1129-1139

 

AREA BIASED NEW-TWO PARAMETER SUJATHA DISTRIBUTION AND ITS APPLICATION TO MEDICAL SCIENCES

 

Dr. V. Arulmozhi, Mrs. D. Narmadha, Dr. R. Sivaranjani, Mr. B. Balachandar

 

This study proposes an Area Biased new two-parameter distribution based on the Sujatha distribution of the relief times patient data. A proposed distribution's characteristics, such as its entropies, moments, moment generating function, survival function, hazard function, and probability density function, are examined. The behavior of a hazard function and probability density function has been depicted using graphs. The well-known maximum likelihood outcome approach is used to estimate the distribution's parameter. Finally, a comparative analysis is carried out to show the flexibility and superiority of this distribution over other distributions. It demonstrates its importance in reliability engineering by generating a rich class of distributions that can better fit empirical data by capturing a variety of shapes and behaviors. The application of this unique distribution is demonstrated using relief times (In minutes) of twenty analgesic data sets. Furthermore, we compare the performance of our distribution with some other existing distributions.

 

Cite: Dr. V. Arulmozhi, Mrs. D. Narmadha, Dr. R. Sivaranjani, Mr. B. Balachandar AREA BIASED NEW-TWO PARAMETER SUJATHA DISTRIBUTION AND ITS APPLICATION TO MEDICAL SCIENCES. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 1140-1149, DOI: https://doi.org/10.24412/1932-2321-2025-489-1140-1149

  

OPTIMIZATION OF INDUSTRIAL SYSTEM MAINTENANCE USING THE MIXED WEIBULL DISTRIBUTION

 

R. Zellagui, L. Khammar

 

This work is part of an approach to optimizing industrial maintenance in a modern and automated production context. It focuses on the in-depth study of the maintenance system implemented in the company IRIS TYRES, specialized in the manufacturing of tires in Algeria, more precisely in the production unit located in Sétif. The main objective of this work is to evaluate the performance of the maintenance applied in the company, based on the fundamental concepts of Reliability, Maintainability and Availability (RAM), as well as modern analytical tools such as the ARROW NOVI CMMS software, the Pareto method. In this article, the Weibull mixture model is proposed to model the failure times of an electrical system in operation with different failure modes and analyze reliability. The estimation of the parameters of the reliability law is done by the maximum likelihood method. Based on IRIS TYRES' system lifetime data, we analyzed and compared the performance of our model with the classic two-parameter Weibull distribution. Strictly speaking, it is necessary to verify which of the two models correctly fits the data. The results of this evaluation show that the machine with a very high failure rate was the curing press, its subsystems (3-bar sensors, valves), and led to the development of action plans to improve equipment reliability, reduce unplanned downtime, and strengthen the efficiency of the maintenance department. The results show that the Weibull mixture distribution tends to overestimate the reliability of the sensors and therefore to overestimate the frequency of preventive maintenance which reduces maintenance costs and the probability of failure.

 

Cite: R. Zellagui, L. Khammar OPTIMIZATION OF INDUSTRIAL SYSTEM MAINTENANCE USING THE MIXED WEIBULL DISTRIBUTION. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 1150-1158, DOI: https://doi.org/10.24412/1932-2321-2025-489-1150-1158

 

IMPROVING THE EFFICIENCY OF ASSESSING THE TECHNICAL CONDITION OF ELECTRICAL EQUIPMENT

 

Muradaliyev A.Z., Ismailova S.M.

 

In recent years, the understanding of the efficiency of technical diagnostics of electrical equipment has increased significantly. The main reasons for this are, in particular: more than half of the equipment in operation has exhausted its estimated service life; the rate of aging of the equipment exceeds the rate of its renewal; due to limited funds for technical maintenance, as well as due to the transition to repairs based on technical condition. High quality and timely diagnostics improve the objectivity of assessing the technical condition of high-voltage electrical equipment. All this allows planning and timely execution of necessary maintenance and repair operations, which makes the operation of electrical equipment more reliable, extends its service life, saves money, and reduces the risk of damage. A method for assessing the moment of occurrence of a change in the growth of concentration of gases dissolved in oil has been developed, allowing to establish the presence of a developing transformer defect, the main gases, the probable type of defect and the nature of its manifestation. A method and algorithm for predicting the time of occurrence of a faulty state of power transformers based on the data of chromatographic analysis of gases dissolved in transformer oil are proposed. A method developed to assess the moment of occurrence of changes in the concentration of dissolved gases in oil, which makes it possible to establish the presence of a developing transformer defect, the main gases, and the likely type of defect and the nature of its manifestation. A method and algorithm for predicting the time of occurrence of a faulty state of power transformers based on chromatographic analysis of gases dissolved in transformer oil are proposed. The recommended algorithm will make it possible to increase the objectivity of comparing the same type of indicators of gases dissolved in oil at the initial stage of wear change.

 

Cite: Muradaliyev A.Z., Ismailova S.M. IMPROVING THE EFFICIENCY OF ASSESSING THE TECHNICAL CONDITION OF ELECTRICAL EQUIPMENT. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 1159-1166, DOI: https://doi.org/10.24412/1932-2321-2025-489-1159-1166

 

OPTIMIZING STEEL INDUSTRY PRODUCTION INVENTORY SYSTEM CONSIDERING WAREHOUSE BOTTLENECK AND MACHINE DOWNTIME

 

Narendra Kumar, Kartik Pandey, Vikash, Kapil Dave

 

The steel industry faces increasing pressure to optimize production and inventory systems while managing operational constraints such as warehouse bottlenecks, machine downtime, and shortages. Addressing these challenges is essential for achieving sustainable cost efficiency and resilience. This study develops a mathematical production-inventory model for the steel industry that incorporates warehouse capacity limitations, equipment downtime, deterioration, and shortage backlogging. The demand is assumed to be time and price-dependent, while production is constrained by effective capacity under downtime. The objective is to minimize the total system cost, which includes production, holding, shortage, deterioration, and downtime penalty costs. The optimization problem is solved using a Newton-Raphson based iterative approach to determine the optimal cycle length and production lot size. Two distinct scenarios are

examined: one with deterioration and one without. Numerical illustrations and sensitivity analyses reveal that warehouse bottlenecks, downtime risks, and shortage policies significantly affect cost efficiency. The results show that downtime reduction and efficient backlog management can considerably lower annual costs. The proposed framework provides practical insights for steel manufacturers aiming to balance production efficiency with inventory resilience, guiding decision-makers in developing sustainable and cost-effective supply chain policies.

 

Cite: Narendra Kumar, Kartik Pandey, Vikash, Kapil Dave OPTIMIZING STEEL INDUSTRY PRODUCTION INVENTORY SYSTEM CONSIDERING WAREHOUSE BOTTLENECK AND MACHINE DOWNTIME. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 1167-1177, DOI: https://doi.org/10.24412/1932-2321-2025-489-1167-1177

 

FORECASTING EXTREME MARKET MOVEMENTS USING REALIZED GARCH AND EXTREME VALUE THEORY MODELS

 

Umme Hani

 

This study investigates the essential benefits of using the Conditional Value-at-Risk (CVaR) optimization framework with the GARCH model and Extreme Value Theory (EVT) methods for the S&P CNX NIFTY markets. Using daily data from Jan. 2020 to July 2024, we evaluate 10 forecasting models, comprising eight standalone GARCH models and eight two-stage GARCH-EVT models. Our results show that the GARCH-EVT models consistently produce more accurate quantile forecasts than their standalone GARCH counterparts. Among the standalone models, daily returns-based GARCH and EGARCH models exhibit the poorest forecasting performance, while the intraday return-based realized GARCH model performs slightly better. The experimental results show that, under various loss functions, the GARCH-EVT innovation model is the best model for volatility predictions of S&P CNX NIFTY among the sixteen forecasting models.

 

Cite: Umme Hani FORECASTING EXTREME MARKET MOVEMENTS USING REALIZED GARCH AND EXTREME VALUE THEORY MODELS. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 1178-1188, DOI: https://doi.org/10.24412/1932-2321-2025-489-1178-1188

 

QUALITY CONTROL FOR SUSTAINABILITY IN CASTING MANUFACTURING

 

Manzoor A. Khanday, Trushal Hirani, Dilawar A. Bhat, Vaddi Tejasvi, Anusha Ale Saji

 

This study proposes an artificial intelligence-driven framework for automating defect detection in casting manufacturing, with a particular focus on enhancing sustainability in the production of submersible pump impellers. A dataset comprising more than 7,000 grayscale images—classified as defective or acceptable—was utilized to train and evaluate three deep learning models: a custom-built Convolutional Neural Network (CNN), VGG16, and ResNet18. The models were assessed for their accuracy, precision, and recall, with VGG16 and ResNet18 demonstrating superior performance, achieving over 99% classification accuracy. Advanced image preprocessing techniques, including histogram equalization and Canny edge detection, were employed to enhance defect visibility and enable the detection of microstructural flaws such as blow holes, shrinkage cavities, and surface irregularities. The integration of transfer learning significantly reduced training time while improving generalization across unseen samples. From a sustainability perspective, the adoption of AI-based automated inspection reduces material wastage, rework, and energy consumption by ensuring early identification of defective products. Furthermore, the deployment of these models on edge devices and integration with IoT-enabled monitoring systems facilitate real-time, continuous quality assurance without human intervention. Overall, the findings underscore that AI-driven quality control systems not only optimize manufacturing efficiency and product reliability but also contribute to achieving broader sustainability goals by minimizing resource depletion, enhancing energy efficiency, and supporting circular manufacturing practices.

 

Cite: Manzoor A. Khanday, Trushal Hirani, Dilawar A. Bhat, Vaddi Tejasvi, Anusha Ale Saji QUALITY CONTROL FOR SUSTAINABILITY IN CASTING MANUFACTURING. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 1189-1200, DOI: https://doi.org/10.24412/1932-2321-2025-489-1189-1200

 

IN COMPLEX INTERVAL-VALUED FUZZY METRIC SPACE, COMMON FIXED-POINT THEOREMS CLR/CLR/E.A PROPERTY

 

Umashankar Singh, Naval Singh, Heera Ahirwar

 

The work of D. Singh et al. served as inspiration for this study, which established complex interval valued fuzzy metric space and proved several common fixed print theorems for compatible types of (p-1) on the structure of complex interval valued fuzzy metric space. The equivalent findings in the body of existing research are expanded upon and generalized by these findings. Additionally, we create a novel implicit relation that can aid researchers in accelerating common fixed-point existence and uniqueness wherever a pair of self-mappings fits the Chr/Ich/E.A-Property. Appropriate examples are provided to demonstrate the usefulness of the primary conclusions that are gained.

 

Cite: Umashankar Singh, Naval Singh, Heera Ahirwar IN COMPLEX INTERVAL-VALUED FUZZY METRIC SPACE, COMMON FIXED-POINT THEOREMS CLR/CLR/E.A PROPERTY. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 1201-1212, DOI: https://doi.org/10.24412/1932-2321-2025-489-1201-1212

 

PROBABILISTIC ANALYSIS OF THERMAL LASER WEAPON SUBSYSTEM IN ANTI-DRONE SYSTEMS

 

Diarmaraja Selvamuthu, Utkarsh Dubey, Raina Raj, Priyanka Kalita

 

The rise of drones has introduced significant security threats, including unauthorized surveillance and explosive payloads. Drones threaten armies by enabling enemy surveillance, facilitating precision strikes on troops or equipment, disrupting communications, and introducing challenges in air defense, thereby altering traditional battlefield dynamics and strategies. To counter these threats, high-energy laser weapon subsystems provide precise, scalable responses with minimal collateral damage. This paper presents a probabilistic analysis of thermal laser subsystem's effects on targets, accounting for aiming and tracking errors. Atmospheric turbulence is modeled using the power spectrum inversion method, while Monte Carlo simulations calculate the on-target spot area. The energy distribution is converted to temperature using thermal physics, and the damage probability is determined by integrating temperatures exceeding material-specific thresholds. Simulation results confirm the system's effectiveness and reliability of thermal laser weapon systems in neutralizing aerial threats.

 

Cite: Diarmaraja Selvamuthu, Utkarsh Dubey, Raina Raj, Priyanka Kalita PROBABILISTIC ANALYSIS OF THERMAL LASER WEAPON SUBSYSTEM IN ANTI-DRONE SYSTEMS. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 1262-1277, DOI: https://doi.org/10.24412/1932-2321-2025-489-1262-1277

 

RELIABILITY, SAFETY, AND SUSTAINABILITY IN NUCLEAR POWER PLANTS: ADVANCES IN SUPERCRITICAL WATER REACTORS AND CIRCULATION LOOPS

 

Pardeep Kumar, Santosh Kumar Rai, Anikate Gupta, Dinesh Kumar, Mahesh Kumar Gupta, Dhowmya Bhatt, Mahiti Gupta, Gyanendra Prasad Bagri

 

In the last two decades, population and energy requirements have drastically risen. Therefore, it is necessary to provide an adequate amount of energy to the people and industries without affecting global warming and the environment. So, several nuclear power plants are constructed to supply the required demand of the power supply to the household and industries. Nuclear power plants can produce more electricity while maintaining global warming norms. However, during production of higher power from the plants, there are several issues raised, such as the sustainability and safety of the nuclear power plant as well as the reliability of the plants. To resolve

these issues to achieve higher plant efficiency, safety of the plant, etc., during operations, several power plants are constructed, such as boiling water reactors and supercritical water reactors to generate power at higher efficiency, while single, two phase and supercritical based natural circulation loops are used to improve the safety of the plant as well as used in operation to improve the power generation from the plant. In past decade, supercritical water reactors and supercritical natural circulation loops have been utilized to increase power generation and enhance the safety of the plant. This review article discusses numerical and experimental investigations of supercritical water reactors and natural circulation loops aimed at enhancing the safety, sustainability, and reliability of various nuclear reactors based on previously published works by several authors, along with identified research gaps and future work scopes.

 

Cite: Pardeep Kumar, Santosh Kumar Rai, Anikate Gupta, Dinesh Kumar, Mahesh Kumar Gupta, Dhowmya Bhatt, Mahiti Gupta, Gyanendra Prasad Bagri RELIABILITY, SAFETY, AND SUSTAINABILITY IN NUCLEAR POWER PLANTS: ADVANCES IN SUPERCRITICAL WATER REACTORS AND CIRCULATION LOOPS. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 1229-1240, DOI: https://doi.org/10.24412/1932-2321-2025-489-1229-1240

 

OPTIMIZATION AND PERFORMANCE EVALUATION OF SUGAR MILL PLANT MICRO COMPONENTS USING PSO

 

Sandeep Kumar

 

Sugar mill plants, especially their core processing units, are highly dependent on a well-organized and efficient maintenance strategy for reliable operation. In this study, an availability-based simulation model was developed for a 25,500 TCD sugar mill. This model is based on a Markov-based performance evaluation framework, which uses differential equations to form state-transition diagrams representing transitions between fully operational, partially operational, and failure states. A comparative analysis of the subcomponents studied concluded that a centrifugal pump failure has minimal impact on overall system availability, while a clarification unit failure has the most significant impact. To improve system reliability and performance, the Particle Swarm Optimization (PSO) method was applied to optimize failure and repair rates. The experimental results obtained show that the overall availability of the system after applying the proposed optimization strategy is achieved at 88.57%, which shows effectiveness and practical utility of the model.

 

Cite: Sandeep Kumar OPTIMIZATION AND PERFORMANCE EVALUATION OF SUGAR MILL PLANT MICRO COMPONENTS USING PSO. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 1241-1253, DOI: https://doi.org/10.24412/1932-2321-2025-489-1241-1253

 

A SYSTEMATIC REVIEW OF MODELING, PLANNING, AND STATISTICAL FRAMEWORKS IN ACCELERATED LIFE TESTING

 

Mustafa Kamal

 

For highly reliable or long-life systems, conventional testing is often impractical due to the excessive time and cost required. Accelerated Life Testing (ALT) offers a practical and statistically rigorous alternative. It predicts product life and durability under normal-use conditions through data obtained under controlled high-stress conditions. This review systematically examines existing ALT methodologies. It begins with an overview of qualitative and quantitative approaches and key concepts that define the theoretical foundation of ALT assessment. Quantitative ALT is organized into Fully Accelerated Life Testing (FALT) and Partially Accelerated Life Testing (PALT) frameworks. Further, the design principles of ALT and PALT are discussed under constant- and

step-stress conditions with emphasis on model formulation, estimation procedures, extrapolation, and model validation. The review also provides a concise discussion of statistical modeling and estimation methodologies. Key elements such as stress-loading, life–stress relationships, and censoring mechanisms are discussed to ensure valid statistical inference and model robustness. The review consolidates theoretical insights and methodological frameworks to advance design optimization and reliability prediction in complex systems. It concludes with perspectives on intelligent, data-driven ALT supported by artificial intelligence and computational analytics.

 

Cite: Mustafa Kamal A SYSTEMATIC REVIEW OF MODELING, PLANNING, AND STATISTICAL FRAMEWORKS IN ACCELERATED LIFE TESTING. Reliability: Theory & Applications. 2025, December 4(89), Vol. 20: 1254-1268, DOI: https://doi.org/10.24412/1932-2321-2025-489-1254-1268