Abstract: At present，the hybridization of Differential Evolution（DE） with local search method confines to DE with a crossover based local search method.A novel differential evolution algorithm with best-individual based Local Search strategy（LSDE） is proposed.In LSDE，a normal distribution operator is introduced to automatically modify search length，and time-varying factors are introduced to modify the two parameters of DE.The experiment results show that except for one function，the ability of finding the optimal solutions using LSDE is better than that of using DE and differential evolution algorithm based on chaos searching （CDE），and the convergence speed of LSDE is quicker than that of DE.

Key words: Differential Evolution（DE）, local search strategy, best-individual

摘要： 针对目前差分进化与局部搜索相结合仅局限于基于交叉的局部搜索的方法，提出了一种基于最佳个体局部搜索策略的差分进化算法（LSDE），并引入正态分布算子自动调整搜索步长和时变差分进化因子调整DE的两个参数。实验结果表明：除一个函数外，LSDE的寻优效果比DE和基于混沌搜索的微分进化算法（CDE）都要好，LSDE的收敛速度比DE快。

关键词: 差分进化, 局部搜索策略, 最佳个体