一种并行遗传算法在家纺企业车间调度的应用  

Application of a New Parallel Genetic Algorithm to Job-Shop Scheduling Problem in Textile Enterprises

作　　者：何桂霞[1] 赵端阳[1] 王雨顺[2] 

机构地区：[1]浙江工业大学之江学院,浙江杭州310024 [2]南京师范大学数学与计算机科学学院,江苏南京210097

出　　处：《工业工程与管理》2008年第2期86-90,共5页Industrial Engineering and Management

基　　金：国家自然科学基金(40405019);国家“973”项目(2005cb321703)

摘　　要：针对家纺企业受特殊工艺约束的车间调度问题,提出了一个基于向量组编码的新的遗传算法,并设计了EOX交叉和启发式变异方法。在基于遗传算法自然并行性特点的基础上,实现了主从式控制网络模式下并行遗传算法。通过仿真实验证明,建立的算法是有效的,收敛速度快,具有较高的并行性,优于普通的遗传算法。Aimed at job-shop scheduling problem with special process constraint in a kind of textile enterprises, a new genetic algorithm which is based on a vector group coding method is proposed, and EOX crossover method and heuristic mutation method are designed. Under the mode of master-slave control networks, parallel hybrid genetic algorithm is applied. The computational results show that it is effective, and is advantageous over common genetic algorithms, and has much better parallel characteristics, and a much better prospect of application can be optimistically expected.

关 键 词：提前/拖后 并行遗传算法 家纺企业 车间调度问题 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]