Prediction Of Minimum Time Path Traveling Based On Urban-grid Vehicle License Plate-Recognition System

The traffic demand grows faster and faster as the socio-economic development recently. Traffic issue can not be dealt well via building roads and facilities simply. The best way to solve this problem is obviously to develop the intelligent transport systems (ITS) in large scales. With the pace of urbanization go so fast, some cities have already implement the project of "Urban Grid", which brings the application of ITS to a new stage. The motivation of this paper is just based on the background of the project.The periodic traffic flow waveform can be found through analyzing a lot of traffic flow data in city grid vehicle license plate-recognition system, especially the weekly period pattern. Using this cycle attribute and improving the model of non-parameter regression, a prediction model based on the similar data in history has been established.To forecast the shortest and second shortest travel time path in city traffic network, using travel time predication combined with the searching of shortest path in this paper, the effective prediction has been obtained and the simulation results have demonstrated its effectiveness.