Abstract:

In order to overcome the shortcomings of the traditional vehicle routing optimization study in terms ofcustomers' commodity demand characteristics,a clustering analysis- based routing optimization using multiple cus-tomer characteristics is proposed.In the investigation,first,linguistic variables are represented by trapezoidal fuzzynumber to implement a comprehensive evaluation of both customers and sub- criterion indices.Next,the sub- criteri-on indices are integrated into a major criterion index via fuzzy integration,and the integrated major criterion value issplit into four sub- criterion values for clustering operation,with a clustering validity index being designed to choosereasonable clustering results.Then,the fuzzy TOPSIS method is used to calculate the customer priority weights foreach cluster.Moreover,evaluation functions for selected customer services are established and are combined withthe dynamic programming method for vehicle routing optimization.Finally,the effectiveness of the proposed methodis verified through an example,and is compared with the existing methods.The results show that the proposedmethod is superior to the method only based on distance measure or customer priority weights,and that it helps toobtain reasonable vehicle routing even in the presence of large- scale customers.

Key words: vehicle routing optimization, customer characteristic, fuzzy clustering algorithm, trapezoidal fuzzynumber