Optimization of Reliability Indices of Repairable Systems Using RPGT and Nature-Inspired Algorithms

Abstract

Reliability of manufacturing system is a crucial determinant of their efficiency, safety and long-term availability. This study develops a comprehensive reliability model of a repairable system using the Regenerative Point Graphical Technique (RPGT). The system’s stochastic behavior is represented through state transition diagrams, from which governing equations are derived for evaluating reliability indices such as Mean Time to System Failure (MTSF), Steady-State Availability (A₀), and the Expected Number of Inspections (V₀). These measures are critical for assessing system performance under varying failure and repair rates. Further, To obtain optimal parameter values and improve system efficiency, three population-based metaheuristic techniques—Genetic Algorithm (GA), Particle Swarm Optimization (PSO), and Cuckoo Search Algorithm (CSA)—are applied. The optimization results are compared to highlight the relative strengths of each method. The findings indicate that CSA achieves better availability outcomes, while GA and PSO provide competitive results for MTSF and inspection measures. The proposed methodology demonstrates the potential of integrating RPGT-based reliability modeling with nature-inspired optimization algorithms to achieve improved performance evaluation and decision-making in repairable systems. This framework can be effectively extended to complex industrial and engineering applications where reliability and maintainability are critical.