基于遗传粒子群混合的可重入生产调度优化  

Scheduling Optimization in Re-entrant Lines Based on a GA and PSO Hybrid Algorithm

作　　者：刘小华[1] 林杰[1] 邓可[1] 

机构地区：[1]同济大学经济与管理学院,上海200092

出　　处：《同济大学学报（自然科学版）》2011年第5期726-730,772,共6页Journal of Tongji University:Natural Science

基　　金：国家自然科学基金重点项目(70531020);国家"八六三"高技术研究发展计划(2007AA04Z151)

摘　　要：可重入生产调度优化问题是个NP难问题,针对可重入生产调度的特点,对该优化问题进行数学规划建模,并通过一些定义将模型映射为有向图,以便于智能搜索算法的应用.结合粒子群算法收敛速度快与遗传算法全局搜索能力强的特点,进行优势互补,并优化设计相关参数,构造了一种混合算法.运用混合算法对供应链优化调度问题模型进行求解,与标准遗传算法、粒子群算法的求解结果进行比较,结果表明混合算法有着更好的优化性能.Scheduling optimization in re-entrant lines proves to be more difficult than in other manufacturing systems,which is well known as a NP-hard problem.A mathematical programming model was established and corresponded with a digraph via several definitions for an intelligent algorithm.A hybrid algorithm was proposed to optimize the objective function,which took the advantages of genetic algorithm and particle swarm algorithm,so this hybrid algorithm integrated global searching ability with high convergence speed.Compared to the results of the normal GA or PSO,simulation results show that the hybrid algorithm is an effective method for scheduling optimization.

关 键 词：可重入制造 调度优化 混合算法 遗传算法 粒子群算法 

分 类 号：F273[经济管理—企业管理]