An Optimization Algorithm Based On Text Clustering And Correlation Analysis For Warehouse Storage Location Allocation

In recent years,with the rapid development of e-commerce and logistics industry,more and more people buy commodities from the Internet,resulting in a sharp increase in online orders which brings great burden to the operation of the warehouse.Therefore many e-commerce enterprises hope to improve the efficiency of warehouse operation to meet the increasing orders.Order picking is a time-consuming process in warehouse operation.Optimizing warehouse location is a method with low cost and high profit to improve order picking efficiency.The storage location allocation strategy adopted by most enterprises is empirical strategy,which has limited effect on the picking efficiency.In the case of few types of goods and strong order correlation,some researchers use the method of correlation analysis to optimize the allocation of warehouse location.But in the case of many kinds of goods,this method is difficult to analyze the relationship between goods.This paper proposes an algorithm based on text clustering and association analysis.In the case of many kinds of goods and weak correlation between items,the algorithm can optimize the allocation of warehouse storage location to shorten the total distance of order picking operation.The work of this paper mainly includes the following points:(1)According to the characteristics of orders,a clustering algorithm suitable for commodity orders is designed.The clustering algorithm aims to improve the correlation between commodities.The clustering algorithm adjusts the number of clustering results by controlling the clustering interval,and divides the sample into different spaces depending on the dimensions of the scaled sample.The clustering interval is the main factor affecting the performance of the algorithm in this paper.(2)Perform association analysis based on the clustering results,and calculate information entropy on the clustering results based on the support obtained by the association analysis.Information entropy can be used as a reference to determine the quality of clustering results.The algorithm in this paper adjusts the clustering interval according to the change of information entropy.(3)Design the calculation method for each type of score,and allocate commodity storage locations based on the score.The scores designed in this paper can reflect the popularity of each type of goods and the association between different types of goods.The principle of allocating locations is that the class with higher score is close to the exit,and the class with lower score is far from the exit.(4)By comparing with other location allocation methods,it is verified that the algorithm in this paper shortens the order picking distance relative to other strategies.In this paper,real business history orders are used in a simulation experiment.Compared with other two commonly used location allocation algorithms,the location allocation strategy proposed in this paper greatly reduces the total distance of order picking operations.Moreover,the storage location allocation strategy in this paper is based on historical orders,which can be adjusted according to the changes of orders,so that the storage location is always in a relatively ideal state,with a strong flexibility.