文献类型：期刊文献

中文题名：基于改进遗传算法的单机成组作业调度

英文题名：Single Machine Scheduling with Job Class Setups Based on Improved Genetic Algorithm

作者：邹律龙[1,2];谭光宇[2];侯东亮[2]

机构：[1]华南理工大学土木与交通学院,广东广州510641;[2]广东海洋大学工程学院,广东湛江524088

年份：2010

卷号：27

期号：4

起止页码：308

中文期刊名：计算机仿真

外文期刊名：Computer Simulation

收录：CSTPCD、、北大核心2008、CSCD_E2011_2012、北大核心、CSCD

语种：中文

中文关键词：优化调度;遗传算法;作业顺序决定的准备时间;总流程时间

外文关键词：Optimal scheduling; Genetic algorithm; Setup - time by job sequence ;Total flow - time

中文摘要：提出了单机成组作业调度的改进遗传算法。优化目标为总流程时间的单机成组作业调度问题明显是NP-hard问题,此问题的多项式求解方法不能求取最优解,而一些启发式算法也只能求出此问题的次优解。为获得单机成组作业最优调度,通过采用整数实值编码,随机采样选择,单点交叉以及变异检查,设计了单机成组作业调度的改进遗传算法。仿真结果表明,算法能够找到此问题的最优解,其性能优于加权最短加工时间(WSPT)启发式算法。改进遗传算法能够灵活解决各种单目标调度及多目标调度问题。

外文摘要：The improved genetic algorithm of single machine scheduling with job class setups is proposed. The problem of single machine scheduling with job class setups which optimal target is the total flow - time is obviously the NP - hard problem. Some polynomial solutions to the problem are not the best ones. And also, some heuristic al- gorithm can only work out the second best solution to this problem. To obtain the optimal single machine scheduling with job class setups, the improved genetic algorithm of single machine scheduling with job class setups is designed by using integer real valued code, stochastic sampling selection, single - point crossover and mutation checking. The simulation result indicates that the improved genetic algorithm not only ean obtain the best optimal solution but also is better than the weighted shortest processing time first （WSPT） heuristic algorithm. By adopting the improved genetic algorithn, various kinds of problems of single - goal and multi - goal scheduling can be solved easily.