Design And Optimization Strategy For Distribution Center Of E-commerce

In recent years, as the popularization of internet, customers can order merchandise in stores by internet, and get them in the next day. Thanks to the convenient operations, e-commerce enters a high speed developing period.In order to support the development of e-commerce, many Chinese enterprises try to develop their own supply chain in e-commerce, for example, Alibaba and Jingdong. The core-competitiveness of a certain supply chain is its whole logistics system, especially the distribution center, because it is responsible for the effective distribution of goods from manufacturers to customers. Compared with traditional sales mode, e-commerce has its own characteristics in logistics, including smaller order sizes, more merchandise types, higher order frequencies, and shorter response times, as customers require orders to be delivered in24hours. In consideration of these new characteristics, the layout of ordinary distribution centers is not very suitable. For example, more merchandise types need too larger storage and picking space that distribution centers can’t afford with ordinary storage and picking operations design. Meanwhile, higher order frequencies and shorter response times require higher order-picking efficiency than traditional sales order. As logistics gradually becomes the bottleneck of e-commerce development, we need to design professional e-commerce distribution center to change this situation.Compared with traditional warehouses, automated storage and retrieval systems (AS/RS) have many benefits, such as increased space utilization and improved material flow. In order to obtain higher space utilization and order processing efficiency, picking operation can be integrated with AS/RS. A common option is an end-of-aisle system, generally called a miniload AS/RS. Because of the increased space utilization and efficiency, the miniload AS/RS is the most available system for e-commerce distribution center among current storage and picking systems.The miniload-workstation order picking system (MWOP) will be used as a basic system to create professional e-commerce storage and picking system, miniload storage and picking system with direct conveyor, which can satisfy all demands of e-commerce distribution center.The main difference between the modified miniload AS/RS and the original one is adding several direct conveyors between corresponding aisles and picking stations in closed-loop conveyor system.During the researching process, the main content and achievement are is as below:(1) For miniload AS/RS, this paper analyzes the order processing and order processing time formulation deeply. Then the main influencing factors of the processing time are proposed. Based on this, the optimization goals and methods are proposed clearly:modify the design of conveyor system and picking stations in miniload AS/RS, and propose the optimization strategies for SKU assignment, order assignment and order sorting problem to fit the improvements of design. With these modifications, the retrieval number of times for SKU will be reduced, and also the distance of retrieval and transportation. Finally, the efficiency of order processing will be improved.(2) With the consideration of characteristics for orders in e-commerce business, the miniload storage and picking system with direct conveyor (MSPD) is proposed, as a modified system for miniload AS/RS. Compared with MWOP, MSPD has three points of improvement. Firstly, add some direct conveyors in original conveyor system of MWOP to reduce the travelling distance of SKU. Secondly, add a buffer area in workstation for picking order boxes to reduce the number of times for retrieval of SKU. Thirdly, add a buffer area in workstation for SKU to reduce the number of times for retrieval of SKU. Analyze the influence of each change for the picking process of system and introduce the whole working process of MSPD. The simulation results showed that, compared with MWOP, the order processing efficiency of MSPD system is improved obviously.(3) SKU and orders assignment. Because there is only a circle conveyor in MWOP, it is not necessary to study the SKU and orders assignment problem. But it becomes necessary to study the SKU and orders assignment problem in MSPD. This paper proposed an optimization strategy to solve the SKU assignment problem. Firstly, the group technology is studied, and then the optimization objective is proposed based on the similarity coefficient. A multi-stage heuristic algorithm is proposed for the SKU assignment problem. The algorithm contains several strategies, such as ABC classification, K-means clustering and local research. Secondly, the orders assignment problem is studies as a clustering problem. Refer to the K-means clustering algorithm, a heuristic algorithm is proposed. The simulation results showed that, these strategies can improve the efficiency of MSPD.(4) Order sorting. Because of the buffer area design in workstation, the order sorting problem is studied in this paper. Firstly, a concept based on sum of the SKU repetitions between several orders is proposed. The model of order sorting is built on the concept. Studied the order sorting problem as a travelling sales problem, the nearest neighbor and nearest insert strategies are proposed. The simulation results showed that, these strategies can improve the efficiency of MSPD, especially the nearest insert one.This paper optimizes the MWOP system, and then improves the order processing efficiency to fit the requirements of e-commerce distribution center. Finally, the professional e-commerce storage and picking system will be built to support the development of e-commerce business.