摘要
提出一种适应性分布式差分进化算法.将初始种群分为多个子种群,并设计子种群间的迁移机制,当满足迁移条件时,根据冯?诺依曼拓扑结构,子种群内的优秀个体代替其邻域的较差个体,使得整个种群实现信息共享.同时,根据个体适应值变化情况,对每一个体分配不同的缩放因子?和交叉率CR,提出?和CR的适应性策略.实验结果表明,所提出算法有利于对解空间进行广泛探索,避免算法陷入早熟收敛,能够搜索到性能较好的解.
An adaptive distributed differential evolution algorithm is proposed based on the change of the individual＇s fitness value. Firstly, the initial population is divided into several subpopulations. When the migration condition is satisfied, the best individual in each subpopulation will replace the worst individual of its neighbor subpopulations according to the Von Neumann topology. The migration mechanism among subpopulations enables the information to be communicated in the whole population. Meanwhile, the adaptive mechanism of F and CR is presented for assigning different F and CR to each individual according to the individual＇s fitness. Numerical results show that, the proposed algorithm is beneficial to explore the solution space, which can avoid the premature convergence and search the excellent solutions.
出处
《控制与决策》
EI
CSCD
北大核心
2014年第4期701-706,共6页
Control and Decision
基金
国家杰出青年科学基金项目(60925011)
国家自然科学基金委国际(地区)合作项目(61120106010)
山西省青年科技研究基金项目(2012021012-4)
太原科技大学校青年基金项目(20113003)
关键词
分布式差分进化
适应性参数
迁移机制
distributed differential evolution
adaptive parameters
migration mechanism
