A Dual Population Collaborative Harmony Search Algorithm with Adaptive Population Size for the System Reliability-Redundancy Allocation Problems

Abstract

Aiming at the problem that the diversity of the current double population algorithm with dynamic population size reduction cannot be guaranteed in real time in iteration and is easy to fall into local optimum, this study presents a dual population collaborative harmony search algorithm with adaptive population size (DPCHS). Firstly, we propose a dual population algorithm framework for improving the algorithm global search capability. Within this framework, the guidance selection strategy and information interaction mechanism are integrated to strengthen the competition and cooperation among populations, and achieving a good balance between exploration and exploitation. A population state assessment method is designed to monitor population changes in real-time for enhancing population real-time self-regulation. Additionally, population size adjustment approach is designed to adopted to effectively streamline population resources and improve population quality. Comprehensive experiment results demonstrate that DPCHS effectively addresses system reliability-redundancy allocation problems with superior performance and robust convergence compared to other HS variants and algorithms from different categories.