文献类型：会议论文

英文题名：Multi-objective hybrid differential evolution algorithm based on Pareto optimal solution migration

作者：Wang, Xiao-Zhen[1]; Li, Peng[2]; Yu, Guo-Yan[2]

机构：[1] Information College, Guangdong Ocean University, Zhanjiang, Guangdong, China; [2] Engineering College, Guangdong Ocean University, Zhanjiang, Guangdong, China

会议论文集：Proceedings - International Conference on Computational Aspects of Social Networks, CASoN'10

会议日期：26 September 2010 through 28 September 2010

会议地点：Taiyuan

语种：英文

外文关键词：Optimal systems - Evolutionary algorithms - Pareto principle

外文摘要：By using differential evolution algorithm (DE) to solve multi-objective optimization problems, Pareto optimal solution migration based differential evolution for multiobjective optimization (PSDEMO) is proposed. The elitist strategy is adopted in the algorithm. Pareto non-dominated solutions found in the evolution operation are archived dynamically with the evolution process, and all the nondominated solutions in the archive are applied to migration operation after mutation and crossover operations of DE are finished. Compared with standard DE, simulation results show the PSDEMO not only helps to improve quantity of the Pareto non-dominated solutions, but also helps to enhance quality of the Pareto non-dominated solutions, it also has good balance keeping ability between diversity and convergence. ? 2010 IEEE.