The SKU Selection Problem In The Self-operated Proactive Warehouses

In order to ensure product quality and improve market competitiveness,fresh e-commerce has derived a proactive warehouse model.In order to reduce delivery time,integrated retail e-commerce has also begun to apply the proactive warehouses model.The competitiveness of the commodities in the warehouse is the competitiveness of the proactive warehouse model.However,the proactive warehouse area is small,and the e-commerce has lots of SKUs.If the area of the proactive warehouse is enlarged,it will decrease the efficiency.Therefore,how to make the optimal SKU selection in a limited area has become a key issue for proactive warehouse operations.Therefore,this article focuses on the SKU selection in the proactive warehouse of self-operated e-commerce.1)Analyzing the connotation of proactive warehouse,it is found that front positions have the following characteristics: integrated storage and distribution model,fast response,and small space.Then summarize the operation characteristics of the proactive warehouse commodity,and find that there is a short replenishment cycle of the product,small daily average shipments and large randomness of shipments.These characteristics determine the daily SKU selection in the proactive warehouse cannot simply rely on the prediction list top N selection.2)Using system theory,starting with the decision-making body,implementation body and feedback body of the SKU selection in the proactive warehouse,combined with their respective goals,the factors that affect the SKU selection of proactive warehouse are analyzed: category Relevance,SKU sales considering substitution effects within categories,SKU gross profit,remaining space in the proactive warehouse,and SKU specification.Then use these influencing factors to build models,use the Apriori algorithm to solve the category confidence parameters,use the LASSO algorithm to solve the SKUs' sales parameters considering the substitution effect,and use the genetic algorithm to optimize.3)Using the data of a fresh food e-commerce company in Beijing,using the Apriori algorithm,LASSO algorithm,and GA algorithm to obtain the optimal plan for the SKU selection of the next day in a proactive warehouse,and comparing it with the plan obtained by the descending sales method,proved that effectiveness of this method.Based on the sensitivity analysis of the results of the case shows that the model has certain robustness.At the same time,suggestions are made on the subjective selection of threshold parameters.