AN ALGORITHM ON MULTI-DEPOTS AND MULTI-TYPE VEHICLES ROUTING BASED ON GLOBAL SEARCHING CLUSTERING

For the multiple vehicle fields multiple vehicle type vehicle routing problem, the paper gives a mathematics model and gives a two stage solving schema. The first stage is using K-means clustering algorithms to change double multi problem to single one to reduce dimension of multiple vehicle fields multiple vehicle model routing problem and reduce the complexity of problem solving .the second stage is using genetic algorithm get shortest routing. Furthermore, the paper give five type interleaved genetic operator and two-step aberrance method to get the initiation population variaty; the schema is proved to be efficient by the experiment.