NSGA-Ⅱ中重复个体的控制  

The Control of Overlapping Solutions in NSGA-Ⅱ Algorithm

作　　者：王广博[1] 韩庆[1] 钟小平[1] 

机构地区：[1]西北工业大学航空学院,西安710072

出　　处：《科学技术与工程》2011年第34期8602-8604,8623,共4页Science Technology and Engineering

摘　　要：由于NSGA-Ⅱ引入了精英策略,为保留父代中优秀个体直接进入子代,在每一代将父代和子代所有个体混合后再选择新种群,所以在新种群中很容易产生重复个体。而进化种群中出现重复个体意味着搜索区域的重叠,使得算法探索新可行区域的效率降低。考虑到计算复杂度和增强解集的有效性,提出了一种删除策略对NSGA-Ⅱ算法进行了改进。算例表明该改进算法能获得分布更加均匀的解集,且具有更强的稳定性。The introduction of elitism strategy of NSGA-II, which can help preventing the loss of good solutions by choosing the best solutions from the merged population that merges the current population and the offspring population to construct the next population, enables easier production of overlapping individuals. The existence of overlapping individuals in the evolution populations means overlapping regions in the searching space, which makes the algorithm much less efficiently in exploiting new feasible region. In consideration of computational complexity and strengthen effectively of solution set. A remove strategy is presented to improve the NSGA-II algorithm. The numerical examples demonstrate that eliminating overlapping solutions make the NSGA-II algorithm more steady and gain a solution set with better distribution.

关 键 词：NSGA—II 重复个体 删除 均匀性 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]