Self-adaptive differential evolution algorithm with improved mutation mode
Self-adaptive differential evolution algorithm with improved mutation mode
复制标题
DOI：
10.1007/s10489-017-0914-3
复制
发表时间：
2017-04
影响因子：
5.3
通讯作者：
Shihao Wang;Yuzhen Li;Hongyu Yang
中科院分区：
文献类型：
--
作者：
Shihao Wang;Yuzhen Li;Hongyu Yang
差分进化算法（DE）的优化性能容易受到其控制参数和变异模式的影响，其设置依赖于具体的优化问题。为此，通过改进差分进化算法的变异模式，引入一种新的控制参数自适应策略，提出了一种改进变异模式的自适应差分进化算法（IMMSADE）。在IMMSADE中，种群中的每个个体都有自己的控制参数，并根据种群多样性和个体差异动态调整。IMMSADE是比较基本DE和其他国家的最先进的DE算法，使用一组22个基准函数。实验结果表明，所提出的IMMSADE的整体性能优于基本DE和其他比较DE算法。
The optimization performance of the Differential Evolution algorithm (DE) is easily affected by its control parameters and mutation modes, and their settings depend on the specific optimization problems. Therefore, a Self-adaptive Differential Evolution algorithm with Improved Mutation Mode (IMMSADE) is proposed by improving the mutation mode of DE and introducing a new control parameters adaptation strategy. In IMMSADE, each individual in the population has its own control parameters, and they would be dynamically adjusted according to the population diversity and individual difference. IMMSADE is compared with the basic DE and the other state-of-the-art DE algorithms by using a set of 22 benchmark functions. The experimental results show that the overall performance of the proposed IMMSADE is better than the basic DE and the other compared DE algorithms.