一种混合智能算法用于求解含保序约束的JSP  

A HYBRID INTELLIGENT ALGORITHM FOR SOLVING JSP WITH ORDER-PRESERVING CONSTRAINTS

作　　者：蒋维[1] 陈开[1] 钟小强[1] 竺长安[1] 王成恩[2] 

机构地区：[1]中国科学技术大学工程科学学院,安徽合肥230027 [2]东北大学教育部流程工业重点实验室,辽宁沈阳110004

出　　处：《计算机应用与软件》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)

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.

关 键 词：车间调度 遗传算法 保序 神经网络 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置] TP393.092[自动化与计算机技术—控制科学与工程]