A stricter constraint dominance principle based algorithm for enhancing multi-performance in constrained multi-objective optimization
提出一种更严格的约束支配原则（SCDP），不再只优先满足可行性，而是同时优化收敛性、多样性和可行性，在32个约束多目标问题中验证了有效性。
The feasibility requirement of constraints poses a severe obstacle to the algorithms’ convergence and diversity when solving constrained multi-objective optimization problems (CMOPs) in science and engineering. This paper adopts a stricter constraint dominance principle, called SCDP, no longer only prioritizing the feasibility optimization while enhancing algorithmic conflicting multi-performance, such as convergence, diversity, and feasibility simultaneously. Firstly, the approach identifies the non-dominated constraints closest to the constrained Pareto Front (CPF) within the feasible regions. Subsequently, convergence, diversity, and feasibility are quantified as competing objectives that stricter the constraint dominance principle (CDP) to optimize multiple performances under non-dominated constraints. The effectiveness of the proposed SCDP is validated through the evaluation of 32 constrained multi-objective problems (CMOPs) and practical applications in the CMOP domain. The results demonstrate that the SCDP based algorithm can improve all the conflict multi-performance of the final solutions when solving CMOPs.