自适应差分演化进化算法

Self-adaptive differential evoultion

针对差分进化算法易出现早熟收敛、局部搜索能力不足及收敛速度慢的特点,提出一种自适应差分演化进化算法(SDE)。该算法在标准差分演化算法的基础上,自适应调整缩放因子(F)和扰动向量,提高算法的搜索速度;混沌调整其遍历方向,促使算法跳出局部极值点,让算法达到全局最优。对多个函数进行仿真试验研究。研究结果表明:该方法具有快速的收敛能力、良好的稳定性,其优化性能有较明显提高。

In order to solve the problems of premature convergence,poor local search and slow covergence speed on differential evolution(DE) algorithm,a self-adaptive differential evolution(SDE) approach was proposed.Based on the differential evolution algorithmm,adaptive adjustment of its scale factor F and disturbance vector were introduced to improve its search speed.In order to make differential evolution jump out of local extreme value point and let the algorithm reach the global optimality,the chaotic adjustment of its traversal direction was added in the approach.The results show that SDE has rapaid convergence,good stablity and better performance than the original DE algorithm.