论文部分内容阅读
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。
Since NSGA-II introduced an elite strategy, in order to retain the elite individuals in the parents directly into the progeny, each generation will be selected after all the individuals of the parent and the offspring are mixed, so it is easy to produce duplicate individuals in the new population. The emergence of duplicate individuals in an evolutionary population implies overlapping of search regions, reducing the efficiency of the algorithm in exploring new feasible regions. Considering the computational complexity and enhancing the validity of the solution set, a deletion strategy is proposed to improve the NSGA-Ⅱ algorithm. The example shows that the improved algorithm can obtain a more uniform solution set and has more stability.