扰动情形下瓶颈利用对作业车间调度的影响

Influence of bottleneck utilization on job shop scheduling under random disturbance

针对瓶颈100%利用后无法应对扰动而造成调度方案无法执行的问题,提出了瓶颈能力释放率和瓶颈能力释放区的概念,研究了瓶颈利用对作业车间调度的影响。在识别瓶颈的基础上,设置瓶颈能力释放率等级;进行瓶颈能力的按级利用,通过遗传算法优化投料次序,输入Plant-Simulation进行模拟仿真,得到各级瓶颈能力释放率的初始调度优化方案;设置机器故障、缓冲、工艺路线更改、交货期变动等随机扰动,对各级初始调度优化方案进行再次模拟仿真,得到随机扰动下各级瓶颈能力释放率的最优调度方案;分析了随机扰动下不同等级的瓶颈能力释放率对作业调度方案的影响。算例结果证实了瓶颈扰动情形下瓶颈100%充分利用并非科学,生成的调度方案鲁棒性差且调度执行结果并非最优的结论。另外,瓶颈非100%有限利用存在最优的瓶颈能力释放率和瓶颈能力释放区,在此释放区之外,过多或过少的瓶颈保护能力对调度方案都有较大的负面影响。

Owing to the 100% utilization of bottleneck machine,job shop scheduling solution could not bear any disturbance which might lead to its failure in practice.To improve the situation,two innovative concepts named Bottleneck Capacity Release Ratio（BCRR） and Bottleneck Capacity Release Interval（BCRI） were proposed,and the influence of bottleneck utilization on job shop scheduling under disturbance was studied.Firstly,on the basis of bottleneck identification,series of BCRR were determined by rank.Secondly,bottleneck capacity was used orderly according to the serial BCRR.By using Genetic Algorithm（GA） and plant-simulation platform,initial optimal scheduling solutions were obtained under different BCRR conditions.Thirdly,simulation environment was modified to simulate real production process in which random disturbances including equipment failure,buffer adjustment,process route change and demand time shift were added.The obtained optimal scheduling solution of each BCRR was inputted to simulate again and then each BCRR＇s optimal solution was obtained.Finally,influence of bottleneck utilization on job shop scheduling under disturbance environment was compared and analyzed.Simulation results demonstrated that robustness and execution performance of job shop scheduling at 100% BCRR were worse than that of non 100% BCRR.When bottleneck was not exploited fully,there existed optimal BCRR and BCRI.However,if BCRR was outside the range of this optimal BCRI,there existed more or less protective capacity of bottleneck resulting in a certain negative impact on the scheduling.