论文部分内容阅读
为有效提高差分进化(DE)算法的优化性能,提出一种动态多子群差分进化(DMSDE)算法.该算法从种群多样性的角度,提出一种动态多子群策略,以增加算法跳出局部极值的可能性.然后,设计了一种平衡局部搜索和全局搜索的随机引导变异操作,以提高搜索的有效性和广泛性.同时,引入全局最优学习操作,防止算法早熟.最后,与差分进化算法和各种改进的差分进化算法及其他智能优化算法做比较,仿真数值结果表明了DMSDE算法的有效性.
In order to effectively improve the performance of differential evolution (DE) algorithm, a dynamic multi-subgroup evolutionary (DMSDE) algorithm is proposed, which proposes a dynamic multi-subgroup strategy from the perspective of population diversity, The possibility of extreme value.And then, a random guided mutation operation is designed to balance the local search and the global search to improve the validity and universality of the search.At the same time, the global optimal learning is introduced to prevent the algorithm prematurely.Finally, Differential evolution algorithm and a variety of improved differential evolution algorithm and other intelligent optimization algorithm for comparison, simulation results show the effectiveness of DMSDE algorithm.