A Multi-Strategy Multi-Objective Evolutionary Algorithm for Solving the Economic Emission Dispatch Problem

The Economic Emission Dispatch (EED) problem seeks to reduce fuel costs and emissions while satisfying load demands and operational constraints. Traditional optimization methods struggle to address such complex multi-objective problems, while evolutionary algorithms with robustness and unsensitive initial conditions can effectively solve them. Therefore, this paper proposes a Multi-Objective Evolutionary Algorithm combining Logistic chaotic mapping, Dynamic Adjustment strategy, and an Adaptive Boosting (AdaBoost) classifier, referred to as LDA-MOEA/D. In the LDA-MOEA/D algorithm, the population is initialized using Logistic chaotic mapping to enhance population diversity and reduce the tendency for premature convergence. Then, a dynamically adjusted mutation factor combining Gaussian and Cauchy distributions was designed to better balance exploration and exploitation, avoiding local optima. Finally, a novel surrogate-assisted selection strategy based on the AdaBoost classifier is proposed to efficiently identify high-quality solutions. Experimental results show that LDA-MOEA/D algorithm performs well on 10 test functions and various practical EED problems.