摘要
车间调度问题是现代制造业快速发展的瓶颈因素,因此提高车间调度的效率和有效性就成了生产制造领域大家普遍关注的问题。现行的车间调度问题遗传算法已经不能满足现代制造业快速发展的要求,多因为其静态性或效率差,表现为早熟或收敛停滞,究其根本原因还是GA的最优参数的选取问题,本文引入了信息熵的概念,以动态调整交叉概率和变异概率,从而给出了可以快速获得最优解的自适应遗传算法,并对此改进算法加以实例仿真验证其有效性。
Job-shop scheduling problem is the bottlenecks in the rapid development of modern manufacture.For this reason,improving the efficiency and effectiveness of the Job-Shop scheduling has been becoming a focus all around the manufacturing field.Being static or poor efficiency,which register as early convergence or stagnation,the existing genetic algorithms used in the job-shop scheduling problems have been unable to meet requirements in the rapid development of modern manufacture.The fundamental reason of this problem is the way that GA select the optimal parameters.To discuss the problem this article introduces a concept of entropy to dynamically adjust the probability of cross and mutation,which makes GA get the optimal solution more quickly.A simulated example is given to show the improved GA's effectiveness.
出处
《中国管理科学》
CSSCI
2008年第S1期142-146,共5页
Chinese Journal of Management Science
基金
863MES项目
广州黄船工程系统项目
关键词
遗传算法
熵
自适应
车间调度
genetic algorithms
entropy
adaptive
job-shop scheduling
