针对非线性多峰函数的优化问题求解困难,提出一种双种群进化策略快速收敛的算法。首先,对于该类最优化问题使用双种群随机变量作为变异算子,在两个不同的子群间并行进行进化,通过使用不同的突变算子策略,实现种群在求解空间具有尽可能分散地搜索的同时在局部也具有尽可能细致的搜索能力。通过子群重组实现子群问的信息交换,通过仿真实例可看出,该算法在非线性多峰值函数优化问题中,具有求解精度较高,收敛速度较快等特点。
For solving the problem of nonlinear muhimodal function optimization, this paper presents a bi - group evolutionary strategies as a fast convergent algorithm. First of all, for this kind of optimization problems, we use bi - group random variables as the mutation operator and the algorithm of evolution in two different subgroups are parallel performed by using different mutation strategic8 to make the group in the solution space to he deeentralizedly explorated as possible while in the local to be carefully done. The exchange of information between supgroups is realized by reorganizing them. We, through the examples, can see that the algorithm in solving the problem of the nonlinear multimodal function optimization has the features like high precision, fast convergence rate etc.