Study On The Optimization Models And Algorithms Of The Feeder Bus Routes For The Urban Rail Transit

The feeder bus is for one kind of public bus for the distribution and collection of passenger flow in urban rail transit, coordinating and cooperating with urban rail transit each other, to form the urban rail transit-feeder bus system. It is of important significance to integrate of urban rail transit and the feeder bus, which both exploit their own advantages, to improve the transportation service of urban passenger transport system. So many deep studies have been done for the feeder bus network design problem(FBNDP). The studies mainly focus on two fields: the study on generation of feeder routes and the study on optimization of feeder bus schedules. The most of the studies on FBNDP are limited to generation of feeder routes without the impact of the schedules of feeder buses. Therefore, in this paper, the method to generate feeder bus routes and the feeder bus schedule optimization method are studied, in order to provide feeder bus and rail transit the effective cohesion theory reference. The main contant of this paper is as fllowing.(1) The characteristics of urban rail transit-feeder bus system are discussed. This paper analyses the technical and economic characteristics of the rail transit and public bus, and the relationship between each other, mutual cooperation between them, respectively exploit their own advantages and characteristics, so as to form effective integration of urban rail transitfeeder bus system. What’s more, passenger flow characteristics and passenger’s travel services characteristics of this system are elabrated in this paper; the different passenger travel cost haulage mode has carried on the quantitative analysis.(2) An algorithm to generate a set of feeder bus candidate lines is designed.Firstly, the model of calculating attract scope based on cluster analysis method is proposed, to determine the scope of the feeder bus service. Secondly main factors that affect the selection of feeder bus stations within attract scope are analyzed. Finally, feeder demand points are given, the method for generating feeder bus candidate lines is proposed, choosing the station with high demand as a feeder line terminal, using the k shortest path algorithm to generate the feeder line between rail stations to each terminal station.(3) The feeder bus network optimization is studied. Firstly, the optimization model with the objective of minimizing the passenger travel time cost and operation cost is established on the basis of analysis of optimization goals and constraints and consideration of the passenger transfer time as an important factor, and feeder routes and bus frequencies are developed using genetic algorithm. Due to the generated lines may exist unreasonable line or station, a heuristic algorithm is proposed to adjust the unreasonable line or site, to generate the final optimization lines. When lines are adjusted, frequencies of the original line are cleared, so the feeder bus schedule optimization model is put forward to produce frequencies, to get feeder bus schedule. Finally the process of feeder bus network optimization is verified by optimization example.The studies on FBNDP have vital significance to the development of urban bus transportation integration. Moreover, it can offer reference for the optimization of urban rail transit and the public bus in certain degree.