MOMS-HDEA: A Multi-Objective Multi-State Hybrid Differential Evolution Algorithm for System Reliability Optimization Design Problems

Abstract

A new custom evolutionary algorithm was developed and implemented to solve multiple objective multi-state reliability optimization design problems. This new algorithm uses the universal moment gene-rating function approach to evaluate the different re-liability or availability indices of the system which have various levels of performance ranging from per-fectly functioning to completely failed. And each com-ponent in sub-system has different performance levels, cost, weight, and reliability. Genetic algorithms are suited for solving reliability design problems because of their appropriate for high-dimension stochastic problems with many nonlinearities or discontinuities. The developed algorithm, MOMS-HDEA, combined the differential evolution algorithm with multi-parent crossover operator satisfying the ergodic and fast properties in searching simultaneously. Experiment also shows that the algorithm gets better Pareto-front solutions.