Abstract:

A flexible job-shop rescheduling of operations with uncertain processing time was studied, and the target was to minimize both the makespan and the total workload of machines. The interval number was utilized to describe the processing time uncertainty and the universal grey number was employed to implement scheduling. Event-driven based on queue size(EDQS) rescheduling triggering mechanism and the rolling horizon optimization were introduced. The conformation method of critical operation set was designed. The complete operation set scheduling was implemented by means of the optimization of critical operation set applied modified NSGA-Ⅱ and heuristic method of non-critical operation set based on shortest processing time. In order to obtain good performance in both convergence and distribution, the hypervolume indicator was introduced to maintain evolutionary population. The order relation of universal grey number was used to ascertain the most optimal scheduling scheme. As demonstrated by simulation, the proposed algorithm which comprises both feasibility and validity is agilely responded the uncertain factors. 

Key words: universal grey number, critical operation set, rolling horizon optimization, rescheduling

摘要：

针对工序加工时间不确定的柔性作业车间重调度问题，以最小化最大完工时间和最小化机器总负荷为目标，利用区间数表征加工时间不确定性，采用泛灰数实施作业调度，引入EDQS重调度触发机制和滚动窗口优化策略，设计了关键工序集的构造规则;采用改进的NSGA-Ⅱ算法优化关键工序集和SPT规则调度非关键工序集，从而实现完全工序集的调度。引入Hypervolume指标维护进化种群，在保证Pareto最优解集收敛性的同时维持了群体多样性，以泛灰数序关系确定最优调度方案。仿真实例不仅证明了基于工序加工时间不确定的柔性作业车间重调度算法的可行性和有效性，而且证明该方法能够及时响应不确定性因素扰动。

关键词: 泛灰数, 关键工序集, 滚动窗口优化, 重调度