Research on a multi-objective constrained optimization evolutionary algorithm
Jiapeng Xiu, Qun He, Zhengqiu Yang, Chen Liu · 2016
To overcome the defects of partial multi-objective constrained optimization evolutionary algorithms especially in getting local optimal solutions, poor diversity and robustness, a hybrid algorithm which is named NCCMOEA (Non-dominated Clonal Constrained Multi-objective Optimization Evolutionary Algorithm) is proposed in this paper. This new algorithm combines the Pareto constrained-dominance, improved stochastic ranking algorithm and clone method in immune multi-objective optimization algorithm. Experiments show that compared with the other effective algorithms, this algorithm NCCMOEA is more excellent in diversity and robustness and avoid getting local optimal solutions obviously.