基于云计算平台的差分进化算法改进研究  

Research on improved differential evolution algorithm based on cloud computing platform

作　　者：孙洁 连畅 SUN Jie;LIAN Chang(College of Electrical Engineering,North China University of Science and Technology,Tangshan 063210,China)

机构地区：[1]华北理工大学电气工程学院,河北唐山063210

出　　处：《现代电子技术》2018年第17期163-166,172,共5页Modern Electronics Technique

基　　金：河北省自然科学基金项目(F2013209326)~~

摘　　要：针对差分进化算法(DE)收敛速度慢、多样性不足和易陷于局部最优的问题,进行基于云计算的差分进化算法改进研究。该方法基于云对差分算法的运行速度进行改善,并以差分算法与轮盘赌相结合的方式改进差分算法易陷于局部最优的问题,即在交叉变异之后对种群以轮盘赌法的方式对其进行筛选,使之能够更好地对个体进行最优的选择,从而实现全局最优的结果。实验结果表明,算法改进后对基本差分算法的收敛速度慢,易陷入局部最优的缺陷有了进一步的改善,并很好地提高了最优解的质量,具有良好的实用性。Since the differential evolution(DE)algorithm has slow convergence speed and poor diversity,and is easy to fall into local optimum,an improved DE algorithm based on cloud computing platform is studied.This method based on cloud platform is used to improve the running speed of the DE algorithm.The DE algorithm and roulette method are combined to improve the problem that the DE algorithm is easy to fall into local optimum,in which the population after cross variation is screened by means of roulette method to select the optimal individual and realize the global optimum.The experimental results show that the improved algorithm can further improve the defects of the basic differential algorithm,greatly improve the quality of the optimal solution,and has high practicability.

关 键 词：差分算法 云计算 轮盘赌 多样性 全局最优 交叉变异 

分 类 号：TN911.1-34[电子电信—通信与信息系统] TP301[电子电信—信息与通信工程]