An Intelligent Hesitation-Weighted Fuzzy Framework For Multi-Objective Reliability Optimization

Abstract

Reliability optimization of engineering systems often involves multiple conflicting objectives and uncertain parameter information, making classical deterministic optimization approaches inadequate. To address this challenge, this study proposes a novel Hesitation-Weighted Sigmoid Membership Function (HWSMF) within an intuitionistic fuzzy optimization framework for solving multi-objective reliability optimization problems under uncertainty. The proposed membership structure explicitly incorporates the hesitation degree of intuitionistic fuzzy sets, enabling adaptive adjustment of satisfaction and rejection levels based on the degree of uncertainty present in system parameters. Component reliabilities are represented using triangular interval numbers, and the resulting interval-valued objectives are transformed into a deterministic intuitionistic fuzzy programming model. The optimization problem is solved using Particle Swarm Optimization (PSO), and the obtained results are compared with those produced by a Genetic Algorithm (GA) to evaluate solution quality and convergence performance. Numerical experiments on benchmark reliability systems, including series systems and space capsule life-support systems, demonstrate that the proposed HWSMF framework provides improved compromise solutions between system reliability and cost. The results indicate that the PSO-based implementation consistently achieves better convergence and higher solution quality than GA. Overall, the proposed approach offers a flexible and robust decision-making framework for multi-objective reliability optimization in uncertain environments, making it suitable for practical engineering system design problems.