Uncertain Information Vehicle Routing Problem Based On Case-based Reasoning

In recent years, with the rapid development of e-commerce, the importance of logistics enterprises has become more and more prominent.Currently more than50%of the total cost is logistics transport cost,the vehicle routing problem can effectively improve the transportation costs,so the research of vehicle routing problem is particularly important.Uncertain events are prevalent in the logistics distribution process,such as sudden vehicle breakdown,congestion, changes in customer demand,these events make the original delivery route uneconomical or infeasible. How to adjust to the original distribution programs fast and making the distribution program still optimal in the unexpected situations is important in the current VRP research,In order to solve the above problems, this paper introduces the theory of artificial intelligence and knowledge engineering to the uncertain information VRP problem solving. In accordance with the thinking of "knowledge representation of VRP problemâ†'build the case baseâ†'case retrievedâ†'case knowledge reuse",this paper solve the VRP problem in unexpected situations from the angle of knowledge reuse based on the case-based reasoning.This can improve the intelligent and real-time of the VRP problem solving.Firstly,this paper combs the theory VRP problem and knowledge representation,proposes a tree structure knowledge representation method of VRP problem.This paper designs the architecture of the support system of VRP problem knowledge representation.This paper aids non-logistics professionals input logistics information by editing the heuristic input window. It describs the knowledge representation tree of VRP problem by the prolog language,generates the VRP problem knowledge-based information model,it can be used for the subsequent VRP problem solving.This paper constructed a case based on the tree structure knowledge representation of the VRP problem.This paper proposed a weight determination method based on the reusability of attributes,it can calculate the case attribute weights in real time and updated the case attribute weights dynamically.On this basis,this paper proposes a multilayer weighted k-nearest neighbor method, it consideres the multilayer affiliation of VRP problem attributes and improves the problem of traditional algorithms whicn didn’t consider the correlation between the attributes. This paper uses MATLAB programming to realize the two algorithms.Finally, taking chain supermarket as application background, this paper uses the tree structure knowledge representation method of the VRP problem, RA algorithm and MWK algorithm to analyze and calculate the unexpected problems in the process of the distribution. The results of the experiment show that the comprehensive application of the methods proposed in this paper can improve the computational efficiency, the full-rate and the resolution.This paper is the cross and penetration of artificial intelligence, knowledge engineering and logistics portfolio optimization, this paper solves the uncertain information VRP problem from knowledge reuse angle based on the cased-based reasoning,provides research ideas and solutions to improve the real-time and intelligentizing of problem solving.