摘要
微分进化算法具有鲁棒性强、易于使用、控制参数少等优点.但在搜索中存在一定的盲目性,搜索效率不高.本文通过引入局部增强算子,使种群中的部分个体在当前最优个体附近寻优,以加快算法的收敛速度.利用五个标准的优化算法测试函数对改进算法进行了测试,并与动态微分进化算法和微粒群算法进行了比较.算法分析和仿真结果表明,本文提出的改进算法大大提高了算法搜索效率.
The differential evolution algorithm is robust,easy to use,and requires few control parameters. However, as the search of the algorithm is of some blindness, its efficiency is limited. To improve the efficiency of the algorithm, the local enhanced operator is proposed to make some individuals of the population search around the current best individual.Numerical study is carried out using five benchmark functions, and the result is compared with that of dynamic differential evolution and particle swarm optimization. Analysis and simulation results show that the efficiency of modified differential evolution is significantly improved.
出处
《电子学报》
EI
CAS
CSCD
北大核心
2007年第5期849-853,共5页
Acta Electronica Sinica
关键词
全局优化
微分进化算法
局部增强算子
global optimization
differential evolution algorithm
local enhanced operator
