基于分拣机器人零售电商订单动态聚类及仿真

目的提高电商物流配送中心订单的分拣效率和动态响应能力。方法通过分析零售电商订单的多品种、小批量和高时效等特征,考虑分拣机器人动作与载重约束,提出了滚动时窗调度策略和高维稀疏动态聚类算法,并以某大型电商企业配送中心建立仿真实验模型,进行数据对比分析。结果以某大型电商某日高峰时段500个订单进行仿真实验,与固定组批分拣策略进行对比,优化后的分拣策略在机器人平均搬运距离上减少了66.9%,分拣时间降低了23.9%。结论高维稀疏动态聚类策略有效提高了分拣效率,降低了分拣成本,算法方式更加开放灵活,对于电商企业物流业务降本增效具有重要意义。

Dynamic Clustering and Simulation of Retail E-commerce Order Based on Sorting Robot

The work aims to improve the sorting efficiency and dynamic response capability of orders for e-commerce logistics distribution centers. Based on analyzing such factors as the characteristics of retail e-commerce orders with multiple varieties, small batches and high aging, movement and load constraints of sorting robots were taken into consideration; a rolling window scheduling strategy and a high-dimensional sparse dynamic clustering algorithm were proposed, and a simulation experiment model of a large-scale e-commerce distribution center was established for data comparison analysis. The simulation experiment was carried out on 500 orders of an e-commerce enterprise in rush hours. Compared with the fixed batch sorting strategy, the optimized sorting strategy reduced the average moving distance of the robot by 66.9% and the sorting time by 23.9%. The conclusion is that the high-dimensional sparse dynamic clustering strategy effectively improves the sorting efficiency, reduces the sorting cost, and the algorithm is more open and flexible. It is of great significance for the e-commerce enterprise to reduce costs and increase efficiency of logistics business.