一种求作业调度问题的改进遗传算法  

A Kind of Improved Genetic Algorithm for JSP

作　　者：王坚生[1] 李举鹏[1] 

机构地区：[1]兰州交通大学信息与电气工程学院,甘肃兰州730070

出　　处：《兰州交通大学学报》2004年第3期48-52,共5页Journal of Lanzhou Jiaotong University

摘　　要：根据JSP问题的特点设计了较为直观的矩阵编码及其相应的交叉和变异操作,对其选择策略中增加用最佳个体来替换一个致死因子,并提出了一个启发修正算子.通过对两种不同模型下的生产作业调度模型的求解,取得了令人满意的结果.According to the character of JSP,a object matrix coding for its individual and relevant crossover operator as well as mutation operator are designed.The algorithm gives a way to replace the lethal with the best individual for its choice strategy and designs a Heurostics correct operator.Problem of standard genetic algorithm is pointed out and way of solving two different kinds of JSP model is given,then satisfied results are obtained.Way for solving similar problems is also provided.

关 键 词：遗传算法 生产作业调度 排序 NP完全问题 

分 类 号：TP182[自动化与计算机技术—控制理论与控制工程]