摘要: 在进化计算中利用群的记忆性,提出群记忆性算法(PMA)。PMA考虑了群个体的当前多样性较优群和多样性权重w1。在多维优化和固定进化次数的情况下,采用Rastrigrin函数、Griewangk函数和Schwefel函数进行测试,benchmark表明PMA的性能优于混沌惯性权重的粒子群优化算法。
关键词:
进化计算,
群记忆性算法,
多样性权重
Abstract: This paper utilizes population memory and proposes a Population Mnemonic Algorithm(PMA) for evolutionary computation. PMA thinks about population individual current more diverse population and the diversity weight w1. With multi-dimensional optimization and fixed evolutionary generations, this paper tests the algorithm by using Rastrigrin function, Griewangk function, and Schwefel function. benchmark shows that PMA is better than Chaos inertance weight PSO(CPSO) algorithm.
Key words:
evolutionary computation,
Population Mnemonic Algorithm(PMA),
diversity weight
中图分类号:
陶俊波, 蔡德所, 吴彰敦, 段秋华. 用于进化计算的群记忆性算法[J]. 计算机工程, 2010, 36(14): 156-157.
DAO Dun-Bei, CA De-Suo, TUN Zhang-Dui, DUAN Qiu-Hua. Population Mnemonic Algorithm for Evolutionary Computation[J]. Computer Engineering, 2010, 36(14): 156-157.