Research Of Vehicle Scheduling Problem Of Hybrid Service Based On GIS

Hybrid Service Vehicle Scheduling is a Vehicle line planning problem that we make a consideration simultaneously pickup service, delivery service and pickup-delivery service. GIS technology can provide us with getting, managing, operating, analyzing and displaying of the geographic space data. This paper made some research about using GIS to describe route net and designed an assistant route network used to assist making decision of vehicle scheduling. This paper made a modal of Hybrid Service Vehicle Scheduling problem, and then used Genetic Algorithm to solute this modal. This paper mainly did the following work and researches:Firstly, aiming at the requirements between any distribution point and distribution center, this paper designed an assistant route network where we used a complete weighted directed graph to describe topology structure between distribution center and distribution points. For the route network with single and double way restrict and turn restrict, this paper put forward a description method of using net restrict set and described the method of saving the route net on the electronic map of GIS. And finally the paper made an improvement on the min-route algorithm. The improvement made the algorithm efficient to generate the assistant route network.Secondly, this paper made research on time window and hybrid service. Basing on the assistant route network, this paper made a Hybrid Service Vehicle Routing model with single distribution centre, single vehicle type and time window. We used service mark to solute the hybrid service problem and used P-D method to solute pickup-delivery service problem. Among this model, we describe particularly parameter definition, object function and restriction conditions, etc.Thirdly, the paper chose Genetic Algorithm to solute this model. This paper analyzed the advantage and disadvantage of the current coding method and designed a new coding method with using logo of the order with a distribution centre arrangement. This coding method is easy to understand and decoding with the advantage of strong expansibility. This paper used adaptive genetic algorithm to control the cross-rate and mutation-rate in order to prevent constringency earliness and it improved the search precision. Meantime we modified the individuals which didn't satisfy restrict and this made the algorithm improved the speed and capability. At end this paper presented the flow of the algorithm.Finally, this paper designed a prototype system of vehicle scheduling of Hybrid Service Based on GIS. Then the paper described the realization of the key function in detail.