摘要
讨论了一种多目标免疫遗传算法的收敛性和多样性。首先,提出了一种集成免疫思想和遗传算法的多目标优化算法;接着,采用马尔可夫链对算法的收敛性进行了定量分析,证明该算法能以概率1收敛到Pareto最优解集;定性分析了算法的多样性保持策略。最后,结合某柔性车间调度问题的实例,验证了算法的良好收敛性和多样性。
The paper discussed the convergence and diversity of a multi-objectictive immune genetic algorithm （MOIGA） and proposed a multi-objective optimized algorithm that integrates the idea of immunity with the genetic algorithm. Furthermore it used the Markov chain to analyze quantitatively its convergence, proving that the algorithm can converge to the Pareto optimal solution sets whert its probability equals to 1. As for its diversity, the paper analyzed qualitatively the strategies of how to maintain it. Finally its convergence and diversity were demonstrated by an algorithm instance that combines a flexible job shop scheduling problem.
出处
《机械科学与技术》
CSCD
北大核心
2006年第12期1480-1483,共4页
Mechanical Science and Technology for Aerospace Engineering
基金
863计划项目(2003AA411110)
教育部博士点基金项目(20040699025)资助
关键词
多目标免疫遗传算法
收敛性
多样性
马尔可夫链
柔性车间调度问题
flexible job shop scheduling problem
multi-objective immune genetic algorithm
convergence
diversity
Markov chain
