混合进化策略算法及其在函数优化中的应用

自适应进化策略中高斯变异算子容易使进化过程陷入局部最优,出现进化早熟.文中针对上述缺点,引入柯西变异算子和子代距离率方法.在进化前期采用柯西的变异,保证个体能够快速地向全局最优的方向移动;在进化后期采用高斯变异,当个体聚集在全局最优解附近时,以较小的变异步长驱动个体向全局最优解方向移动.子代距离率系数进行调整变异算子.通过对单峰与多峰函数仿真试验,验证了算法的有效性.

Hybrid Evolutionary Strategy Algorithm and Its Application in Function Optimization

In adaptive evolution strategy,Gaussian mutation operator made the evolutionary process fall into local optimal,the evolution appeared premature.Aiming at the shortcoming,introducing Cauchy mutation operator and progeny distance rate method.In the early evolution,using the variation of Cauchy,ensure the individual can quickly move to global optimal direction.In the later evolution using Gaussian mutation,when the individual gathered in the vicinity of global optimum solution,with smaller mutation step drive...