基于单纯形局部搜索的自适应差分进化算法

针对标准的差分进化（DE）算法在高维复杂的函数优化中易早熟收敛，进而导致搜索精度低甚至优化失败的问题，提出一种基于单纯形局部搜索的自适应的差分进化算法（SSADE）．将DE算法的快速全局搜索能力与单纯形的强局部寻优能力有机结合起来，进一步提高了解的精度．参数自适应变化有效地维持了种群的多样性，自适应的变异策略扩大了个体的搜索范围，增强了算法寻优效果，仿真实验验证了新混合算法的有效性．

Self-adaptive Differential Evolution Algorithm Based on the Simplex Local Search

In the report, self-adaptive differential evolution algorithm based on the simplex search （SSADE） was proposed to solve the premature convergence and low precision of standard differential evolution （DE） applied for the optimization of high-dimensional complex function. The SSADE hybrid algorithm organically integrated DE algorithm which has powerful global search capability with the simplex search method which has strong local search ability, and which further improved the precision of solution. The parameter self-adaptive adjustment maintained the diversity of the population, the self-adaptive mutation strategy expanded the search range of the individual, enhanced the effect of the algorithm optimization. The simulation results confirmed the effectiveness of the new hybrid algorithm.