求解二级分销网络模型的混合微粒群算法  

Mixed-particle swarm optimization for two-level distribution network model

作　　者：周鲜成[1] 赵志学[1] 贺彩虹[2] 徐戈[2] 

机构地区：[1]湖南商学院计算机与电子工程学院,湖南长沙410205 [2]湖南商学院会计学院,湖南长沙410205

出　　处：《中南大学学报（自然科学版）》2010年第2期623-627,共5页Journal of Central South University:Science and Technology

基　　金：湖南省科技计划项目(2009GK3054);湖南省教育厅重点科研项目(09A048)

摘　　要：针对现有算法在求解二级分销网络模型时计算量大、难以适用于求解大型规划问题和易陷入局部最优等不足,提出一种求解二级分销网络模型的混合微粒群算法。该算法以二级分销网络的总成本作为适应度函数,采用一种精简的编码方式,通过将遗传算法的变异和交叉操作引入微粒群算法,实现二级分销网络模型的离散优化。算例仿真结果表明:采用提出的算法能获得全局最优解,且收敛性好,运算速度快,稳定性好,能有效避免算法的早熟收敛问题。Based on the fact that there are large calculation, difficulty of solving large programming problems, and possibility of falling into local optimum, a mixed-particle swarm optimization was proposed to solve such disadvantages of two-level distribution network model. The proposed algorithm took the total cost of two-level distribution network as the fitness function, and a streamlined encoding was applied. The particle swarm optimization was combined with the mutation and crossover operations of the genetic algorithm in the proposed algorithm to realize the discrete optimization of two-level distribution network model. The simulation results show that the proposed algorithm can effectively avoid the premature convergence, have the access to the global optimal solution, and can enhance efficiency of algorithm.

关 键 词：二级分销网络 混合微粒群算法 遗传算法 供应链 

分 类 号：TP202.7[自动化与计算机技术—检测技术与自动化装置] F203[自动化与计算机技术—控制科学与工程]