Research On Customized Bus Route Planning Based On Granular Computing

To alleviate urban traffic congestion and environmental pollution for meeting different demand of passengers,a bus service based on demand response is promoted,called customized bus.The route planning of customized bus consists of three parts:travel demand acquisition,stations planning and routes planning.However,there are some problems in traditional planning scheme,such as non-diversification method of needs acquisition and unreasonable planning of stations and routes,which resulted in a low rate of customized bus bookings or even inability to conduct.Therefore,it is necessary to make a in-depth research on the routes planning of customized bus.The main problem with customized bus route planning is stations planning.The stations planned by the location of passengers travel demand.The essence of this problem is a clustering process.Granular computing is a new methodology for dealing with complex problems.Include the thinking of granular analysis in the clustering process can simplify the clustering work and obtain better clustering results.Under the guidance of granular computing theory and traffic big data,this paper optimizes the problems existing in the tradition customized bus route planning scheme,It is difficult to make route planning when travel demand are low.In this paper,the route planning scheme customized bus based on the traditional bus stations and the actual travel demand of passengers is proposed to solve the problem above.The travel demand of the station can be judged by analyzing the passenger traffic of the traditional bus stations.And it can be converted to passenger travel demand for supplementary of total travel demand of customized bus by certain rules to increase the impact of the market on route planning results,which provides a feasible thinking for customized bus route planning in the scenario of less travel demand.In the planning of the station,this paper proposes a station planning method based on the majority of passengers ’ willingness and travel distance.Firstly,quotient space granularity transformation thinking and similarity adjustment factor are introduced on the basis of AP clustering algorithm,which is used to improve the number of clusters obtained and the applicability of the algorithm with its different node weights.In this way,the final result will satisfy the majority of passengers’ willingness and travel distance.The feasibility of the improved algorithm in the route planning problem of the stations is proved by the combination of simulation experiments.In terms of driving route planning,this paper designed the driving route cost calculation model according to the four aspects of the travel cost of passenger,the cost of company operating,the cost of road resource and the cost of environmental pollution,respectively.It is aimed to optimize the update rules of ant colony algorithm pheromone,and to obtain the driving route with relatively low cost and relatively short path.Finally,the simulation based on the data by the Chengdu bus platform is carried out.And the feasibility of this scheme in the customized bus route planning is verified through the experimental evaluation and analysis.