论文部分内容阅读
为了克服遗传算法中存在的主要问题即过早收敛(过早收敛使得一些优秀个体或基因过早地被排除掉,从而导致搜索范围缩小及局部最优,影响了进一步搜索),从控制参数的改进着手,提出了多种群并行进化及自适应调整控制参数相结合的思想。克服了以往定常参数单种群进化的不足,综合了不同特性种群进化的长处,使得过早收敛问题得以缓解,同时又提高了搜索的范围和效率。
In order to overcome the main problem existing in genetic algorithms, ie, premature convergence (premature convergence makes some excellent individuals or genes prematurely excluded, resulting in narrowing of the search range and local optimization, affecting further search), from the control parameters Improve and put forward the idea of multi-population parallel evolution and adaptive adjustment of control parameters. It overcomes the shortcomings of the single population evolution of the constant parameters in the past and integrates the advantages of the population evolution with different characteristics so that the problem of premature convergence can be alleviated and the search scope and efficiency are also improved.