Abstract: Aiming at the problem of high real-time requirements and easy schedule deviation when dynamic events occur,a dynamic scheduling approach to Robotic Mobile Fulfillment System (RMFS) was proposed.Characteristics of dynamic events in RMFS were analyzed.The consequences of different dynamic events and the effects of scheduling were determined.A mathematical model aiming to minimize the total completion time was presented.A dynamic scheduling approach was used to realize pre-scheduling of real-time orders arriving and rescheduling after dynamic events in RMFS.According to the different effects of dynamic events,the rescheduling rules with occurrence of dynamic events were established.The dynamic scheduling approach could promptly respond to dynamic events such as urgent orders and robot failures.Compared with traditional approach,the solution was obtained by dynamic scheduling approach with shorter total completion time and path length.After failure time re-schedule affected autonomous vehicle had less impact on the re-scheduling result.

Key words: robotic mobile fulfillment system, dynamic scheduling, real-time scheduling, rescheduling rules

摘要： 针对多机器人存取系统实时性要求高,动态事件发生时易偏离调度计划的难题,提出一种动态调度方法。对系统内发生的动态事件进行分析,确定不同动态事件造成的后果和对调度的影响。建立以完成时间最小为目标的动态调度模型。提出动态调度方法实现系统内实时订单到达情况下的预调度和动态事件发生后的重调度。根据动态事件的不同影响,确立动态事件发生时的重调度规则。通过实例证实,相比现有调度方法,所提出的动态调度方法能够及时响应紧急订单到达和机器人故障等动态事件的发生,获得完成时间更小且路径长度更短的调度方案,选择在故障时间后对受影响机器人进行重调度对重调度结果的影响更小。

关键词: 多机器人存取系统, 动态调度, 实时调度, 重调度规则