论文部分内容阅读
差分进化算法被认为是一种简单高效的全局优化算法,但其在解决高维复杂优化问题收敛精度过低,为此提出了阶段波动差分进化算法.新算法利用柯西分布随机数设计用于生成变异率的算子,并对变异率进行上下波动.进化过程中引入分段思想,每个阶段分别根据不同的配置利用算子生成变异率并选择对应的交叉比率改善算法性能.同时为了加快收敛速度,设计了一种新的变异处理策略.通过对一组经典Benchmark函数的测试,实验结果显示了本文算法在解决复杂高维问题时具有优于或相当于其他DE算法的性能.
Differential evolution algorithm is considered as a simple and efficient global optimization algorithm, but its accuracy in solving high-dimensional complex optimization problems is too low, so a staged differential evolution algorithm is proposed. The new algorithm uses Cauchy distributional random number design Generate the operator of mutation rate and fluctuate the mutation rate.Segmentation idea is introduced in the evolutionary process and each step is used to generate the mutation rate according to different configurations and select the corresponding cross ratio to improve the performance of the algorithm.At the same time, Convergence rate, a new mutation strategy is designed.Through testing a set of classical Benchmark functions, the experimental results show that the proposed algorithm has better performance than other DE algorithms in solving complex high-dimensional problems.