Searching for Robust Optimal Solutions by an Evolutionary Algorithm

摘要：

When searching for robust optimal solutions in evolutionary optimization, a criterion for robustness estimation is needed. An expectation-based approach and a variance-based approach are highlighted in this paper. Experiments are conducted on a singleobjective test function adopting three unequal values of dimension size, solved by an chaotic evolutionary algorithm. In both former low-dimensional cases, graphical results are given to make a visual comparison between the two robustness measures, which shows that their searching ability is complementary. Then a dominance mechanism similar to the definition of domination in multi-objective optimization is suggested, and integrated with chaotic evolutionary algorithm for selection of a survival between two individuals in single-objective evolutionary optimization. In the third high-dimensional case, solved by modified chaotic evolutionary algorithm with variance-based measure approach, good numerical results are reported.

关键词： robust optimal solution robustness evolutionary algorithm optimization

作者: Minling Wang Xiufen Zou

作者单位: Department of Mathematics & Physics Wuyi University Jiangmen,China School of Mathematics & Statistics Wuhan University Wuhan,China

会议类型: 国际会议

会议名称: The 2nd IEEE International Conference on Advanced Computer Control(第二届先进计算机控制国际会议 ICACC 2010)

会议地点: 沈阳

会议语种:英文

页码: 593-595

在线出版日期: 2010-03-27（万方平台首次上网日期，不代表论文的发表时间）