Two-Populations and Unconstrained Solution Selection for Constrained Multi-objective Optimization

Due to the presence of numerous infeasible regions in Constrained multi-objective optimization problems (CMOPs), existing algorithms struggle to identify the complete Pareto front (PF). A novel two-populations algorithm is introduced to address these challenges. This algorithm includes the improved ∈-method and a new environment selection strategy that incorporates unconstrained solutions. These enhancements facilitate the rapid convergence of the population towards the PF, reducing the negative impact of infeasible regions on the results. In addressing three-dimensional CMOPs, we assess population density using an angle-based niche method, which effectively maintains population diversity. Our method is compared with five state-of-the-art methods using the 2 benchmark test suites and 14 real-world CMOPs. The experimental results demonstrate that the algorithm can identify a more complete PF.