摘要
针对车间调度中计算复杂度问题,提出将神经网络嵌入遗传算法中,在初始化序列时考虑到工件中工序的加工顺序,采用基于保序的方法来对染色体进行交叉和变异。实验仿真表明,该算法能够获得比较理想的加工序列,在指定的代数内能够收敛于优值。
In this paper it presents to embed the neural network into genetic algorithm in light of the computation complexity problem in job-shop scheduling. When initializing the sequence, the job order of processing for work pieces is taken into consideration and the chromosomes are crossed over and mutated in a way based on order-preserving. From the emulation experiments it is illustrated that the algorithm can obtain a rather ideal job order and is able to converge to optimal value in assigned algebra.
出处
《计算机应用与软件》
CSCD
北大核心
2008年第12期70-72,共3页
Computer Applications and Software
基金
中国科学院创新基金资助项目(200417009)
关键词
车间调度
遗传算法
保序
神经网络
Job-shop scheduling problem Genetic algorithm Order-preserving Neural network
