基于RPA与滚动时间窗的零售电商订单动态调度仿真

Dynamic Scheduling Simulation of Retail E-commerce Orders Based on RPA and Rolling Time Window

为了提高零售电商订单的整体交付效率,提出基于RPA与滚动时间窗的零售电商订单动态调度方法。在RPA机器人技术框架的应用层逻辑端中,通过基于滚动时间窗的订单调度优先级排序方法,将订单以滚动时间窗划分的方式,分解为多个订单调度任务,并执行订单优先级排序。将排序后订单作为零售电商订单拣选与配送车辆联合动态调度方法的调度对象,构建订单交付时间最小化、全部订单均被交付的订单动态调度目标函数。由改进共同进化遗传算法,从订单拣选、配送车辆联合动态调度的角度,求解满足条件的订单拣选与配送车辆联合动态调度方案。实验验证,此方法调度下,零售电商客户订单配送交付时,交付时间均靠近客户对订单产品期望时间窗下界,无订单遗漏问题,全部订单均被高效拣选、配送。

In order to improve the overall delivery efficiency of retail e-commerce orders,a dynamic scheduling method of retail e-commerce orders based on RPA and rolling time window is proposed.In the application layer logic end of RPA robot technology framework,this method decomposes the order into multiple order scheduling tasks in the way of rolling time window division through order scheduling priority sorting method based on rolling time window,and performs order priority sorting.Taking the ordered order as the scheduling object of the joint dynamic scheduling method of order picking and distribution vehicles for retail e-commerce,the objective function of order dynamic scheduling with minimum order delivery time and all orders being delivered is constructed.From the perspective of order picking and distribution vehicle joint dynamic scheduling,an improved coevolutionary genetic algorithm is used to solve the joint dynamic scheduling scheme of order picking and distribution vehicles that meet the conditions.Experiments verify that under this method,when the retail e-commerce customer orders are delivered,the delivery time is close to the lower bound of the customer’s expected time window for the order products,and there is no order omission problem.All orders are efficiently picked and delivered.