Optimization Method Of Multi-distribution Centers Location Based On K-means Clustering Algorithm And Evidential Reasoning Approach

Distribution center is considered as the important hub of logistics system,it's location selection problem decides the smoothness,economy and order of the whole enterprise logistics system.For retail enterprises,scientific and rational location selection can reduce the transportation cost,improving the efficiency of the goods,helping enterprises to preempt the market,improving customer service satisfaction,to ensure long-term healthy development of the logistics system.Therefore,this paper is based on the existing distribution center location methods,the clustering algorithm and evaluation of site selection method based on fuzzy mathematics,the K-means clustering algorithm and evidential reasoning method,is used to solve the candidate points have been identified,the number and position uncertainty distribution center location problem."build multi-distribution center in an area of the site selection problem" transfer into "respectively in the area of the single distribution center location problem".First using K-means clustering algorithm,based on Euclidean distance between alternative points similar conditions,cluster the candidate points,take different values for K of clustering results;Second,to determine the number of distribution centers location selection is scientific and reasonable,introduces BWACR clustering validity index,analyze the K-means clustering results,the effect of clustering gets better as the BWACR parameter values increases,and the corresponding K value is the best location number;Then according to the influencing factors of distribution center location selection,a level 4 location comprehensive evaluation index system is established,using the evidential reasoning approach to integrate the quantitative and qualitative factors,using IDS software to calculate sets' alternatives comprehensive evaluation of the utility values,choosing the alternative with highest utility values in set as the best location;Finally,a numerical example is presented to illustrate the validity and the practicability,and can help deciders in real life make more scientific and reasonable decisions.