考虑随机因素的单件车间生产作业计划方法研究

Study On Stochastic Job Shop Scheduling Problem

本文应用改进的随机动态神经网络解决单件车间生产作业计划问题(JSP),然后在以较好的生产作业计划为起点寻优的过程中,使用了基于工件加工时间服从概率分布的禁忌搜索,并且用TurboC++1.0和Foxpro 2.5 for Windows在微机上实现上述神经网络和禁忌搜索方法.

For neural networks approach to the job shop scheduling problem,an improved stochastic dynamical neural network is used and the quality of job shop schedule is improved. In the course form the good job shop schedule to the best job shop schedule, A Tabu search based on job' s completion time described by probability distributions is used. All above have come true on the computer by Turbo C + + 1. 0 and Foxpro 2. 6 for Windows.