摘要
本文讨论了MOGA目前存在的缺陷,并提出利用共享小生境技术为基础更新子群体,并针对遗传迭代过程提出相应的改进遗传策略。策略包括采用了期望、精英保留混合策略以及改进快速自适应的交叉、变异算子。最后,利用改进遗传算法在多目标文献中作实例研究,并取得了良好的应用效果。
This paper discusses MOGA's existing deficiencies,and offers a method to update the next population based on the niche technology. During the genetic iterative process, this paper adopts the corresponding improvement strategies which include expectations, the elite reservation hybrid strategy, and the improved fast adaptive crossover and mutation operator. Finally we use the improved GA to study the multi-object case. Simulation experiments show that NGA can effectively resolve the convergence of the multi-objective optimization problem.
出处
《计算机工程与科学》
CSCD
2008年第3期75-77,共3页
Computer Engineering & Science
关键词
多目标规化
遗传算法
小生境
改进快速自适应
multiobjective optimization
genetic algorithm
niche, improved fast adaptive technology