Reliability optimization of non-linear RRAP with cold standby through HPSOTLBO

Abstract

The cold standby reliability redundancy allocation problems (RRAP) with nonlinear constraints have significant challenges in optimization due to their complexity. This work aims to address these problems by developing a hybrid optimization technique called hybrid particle swarm optimization with teaching–learning-based optimization (HPSOTLBO). In this study, the basic methodology of the proposed research is a metaheuristic approach. Its primary objective is to maximize system reliability by optimizing redundancy and component reliability through the balance between local exploitation and global exploration. The HPSOTLBO algorithm combines the robust global search capability of particle swarm optimization (PSO) with the rapid convergence features of teaching–learning-based optimization (TLBO). The hybrid approach, TLBO dynamically updates the PSO’s searching processes, enhancing its effectiveness in locating optimal solutions. The algorithm is tested on three benchmark problems for reliability optimization with cold standby strategy to demonstrate its practical utility. Computational experiments reveal that HPSOTLBO consistently outperforms both PSO and TLBO individually, as well as several previously utilized metaheuristic approaches. According to outcomes, HPSOTLBO provides a strong framework for nonlinear RRAP. Further, the statistical analysis using the Friedman ranking test and Wilcoxon sign-rank test validates the algorithm’s superior efficiency. Finally, the study shows that HPSOTLBO is an effective technique for resolving RRAP with a cold standby strategy.