基于混合GA算法求解车间调度问题  

Solving job shop scheduling problem based on hybrid GA algorithm

作　　者：王粟[1] 陈新彦 曾亮 WANG Su;CHEN Xin-yan;ZENG Liang(School of Electrical and Electronic Engineering,Hubei University of Technology,Wuhan 430068,China)

机构地区：[1]湖北工业大学电气与电子工程学院,湖北武汉430068

出　　处：《计算机工程与设计》2022年第5期1304-1311,共8页Computer Engineering and Design

基　　金：湖北省重点研发计划基金项目(2020BAB114)。

摘　　要：由于车间调度问题组合排序众多等复杂性因素的存在,使用遗传算法求解时,初始种群的随机产生和变异的随机发生对寻优的效率影响很大。针对上述问题,提出一种混合GA算法,主要从变异策略和种群生成两方面进行改进,采用经过选择、交叉操作种群的平均适应度值来决定是否进行变异操作,借鉴SA算法中的重升温策略,将引入自适应控制因子和排列操作的PSO算法产生的个体极值种群代替GA算法特定代数的种群。仿真结果验证了该算法求解车间调度问题的有效性。Due to the existence of many complex factors such as combinatorial scheduling of job shop scheduling problem,the random generation of initial population and the random occurrence of mutation have great impacts on the efficiency of job shop scheduling.To solve the above problems,a hybrid GA algorithm was proposed,which mainly improved the mutation strategy and population generation.The average fitness value of the selected and cross operated population was used to decide whether to carry out mutation operation or not.Referring to the reheating strategy of SA algorithm,the population of a given algebra using GA was replaced by the individual extremum population generated by PSO algorithm with adaptive control factor and permutation operation.The simulation results verify the effectiveness of the algorithm.

关 键 词：遗传算法 粒子群算法 模拟退火算法 作业车间调度 重升温 

分 类 号：TP301[自动化与计算机技术—计算机系统结构]