鹰策略结合差分进化的可靠性冗余优化算法  

Reliability redundancy optimization algorithm combined eagle strategy with differential evolution

作　　者：刘玉宝[1] 祝海英[2] 李纯莲[1] 戴银飞[1] 

机构地区：[1]长春大学计算机科学技术学院,吉林长春130022 [2]长春职业技术学院信息技术分院,吉林长春130033

出　　处：《辽宁工程技术大学学报（自然科学版）》2015年第6期763-768,共6页Journal of Liaoning Technical University (Natural Science)

基　　金：吉林省教育厅"十二五"科学技术研究项目(吉教科合字[2013]第461号);吉林省科技发展计划项目(20140204048GX)

摘　　要：将鹰策略和差分进化结合用于解决可靠性冗余优化问题.优化过程分为两个阶段:第一阶段使用Lévy飞行在解空间中进行全局搜索,第二阶段使用差分进化算法在前阶段得到的有前途解的周围进行快速的局部搜索.同时,修改了差分进化算法的变异算子和交叉算子以提高局部搜索的性能.该算法较好地实现了全局搜索和局部搜索的平衡,既有利于跳出局部最优,又可以加快局部收敛.通过对可靠性冗余优化的两个基本问题的实验表明,所提出的算法在解决可靠性冗余优化问题上是有效的.Eagle strategy is combined with differential evolution for solving reliability redundancy optimization problem. The optimization process is divided into two phases： First,levy flight is used to do global search in the solution space; Second, differential evolution algorithm is used to do fast local search around a promising solution. At the same time, the scale factor F and crossover rate CR are modified to improve the performance of local search in differential evolution algorithm. The proposed algorithm achieves the balance of the global search and local search. It is not only conducive to jump out of local optimum but also speed up the local convergence. Experimental results over two benchmark problems demonstrate that proposed algorithm is effective for solving the reliability redundancy optimization problem.

关 键 词：鹰策略 自适应差分进化 可靠性优化 冗余分配 非线性规划 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]