Optimization of Storage Allocation of Mobile Racks Based on Demand Forecast

Abstract

The complexity of a single multi-product order structure of retail e-commerce enterprises. There is a difficult situation of time-consuming step-by-step picking and a heavy load of rack handling. The product fails to take advantage of the flexibility of the mobile racks. Matching with the shelves according to the sales rules results in a long pick-and-walk distance. The general classification storage strategy in the warehouse makes it difficult to deal with fragmented picking. Given this, the optimization of the storage position of mobile racks in the warehouse of retail e-commerce enterprises is studied. A combination optimization model is established to minimize the length of the pick road. The first stage is the introduction of complex network abstraction of each commodity for the nodes to establish the correlation network, mining within the order of the commodity demand correlation for clustering; In the second stage, a subset of (1, N) Grey-Prediction model related factor series is used to distribute racks for the first stage of clustered commodity portfolio using single parent evolutionary genetic algorithm under elite strategy based on short-term sales prediction. The evaluation of the experimental data shows that compared to the current storage strategy of the enterprise, the optimized model yields a better-quality solution. Robustness perturbed by order size. At least a 2.85% reduction in pick length at large parameters. Rack handling is reduced by at least 5.62%. The result of a short distance and a small number of picks is realized.