Study On The Arithmetic Of The Road Networks Capacity And Traffic Volume Forecast

Based on the research of road network's capacity at home and aboard ,we make up research on the application of traffic volume's combined-forecast model based on BP neural network and graph theory method in the project of road network from the perspective of road network. Traffic volume's prediction is closely related to the project of road network, and the request that traffic ability satisfies traffic volume in the future is directly related to the work of the project of road network. Especially, the arterial road network carries most part of road network's traffic volume, its capacity can provide basic data and quantitative theoretical basis for the work of transport's project and management of the city. The research of its capacity is an essential for the perfect of the traffic engineering theory.In the first it reviews and analyses history, status in quo and trend of internal and oversea study in capacity of road networks. Concretely researching existed methods of the traffic volume prediction, bringing forward combined-forecast model based on BP neural network, and carrying concrete data and verifying better result compared with single model. It goes into graph theory in many ways, applicability, advantages and disadvantages, then finds out an algorithm—converts the method of solving maximum flow problem to shortest road problem. Adopting the abutment chain method of the data structure to structure assistant graph and make the project of large-scale road networks more convenient and understanding. Subsequently, it applies the method to research urban road networks, and find out the min-cut capacity and the key road segments that supervise network capacity. In the end of the paper I summarized the fruits of my work and I put up suggestions of researching emphases for the future. I hope it have valuable try for the study on the project of road networks.