The work aims to alleviate the problems of large area occupation of retail e-commerce commodity warehouse and limited picking efficiency. The Auto Store warehousing system was introduced in detail in terms of storage strategy, assignment strategy and routing strategy. On this basis, the business process of the Auto Store system to complete a single order picking operation was analyzed. The Anylogic software was used to simulate and verify the proposed model. The picking efficiency of the mixed storage layout and the traditional layout of Auto Store warehousing system was compared and the feasibility of the mixed storage layout was verified when the number of picking vehicles and workstations were the same as the number of picked goods under the assumption that the order arrival obeyed the Erlang distribution. At the same time, according to the comparison based on the two task assignment strategies, it is concluded that the the assignment strategy based on the minimization of picking time has higher system operation efficiency. The results between theoretical analysis and simulation research are analyzed and compared and the effectiveness of the model is verified.