Research On Intelligent Traffic Information Collection And Guidance Strategy

With the continued growth of urban motor vehicles, road congestion has become an important factor in urban development. As a new way to solve the problem of urban traffic, intelligent traffic system(ITS) obtains rapid development in recent years. In order to avoid the occurrence of road congestion and improve the operational efficiency of urban road network, how to develop effective vehicle guidance information becomes very important for traffic managers. At the same time, intelligent traffic information collection system provides basic data to the traffic guidance information, and therefore the data collection process is essential to the development of guidance information. For the above analysis, traffic information acquisition and processing and guidance strategy in ITS is studied in this thesis, mainly as follows.Firstly, this thesis researches a method to judge the parking status of the floating car to sovle the drift data and redundant data in intelligent traffic information collection process. The method judges the parking status by combining the stopping time and the longest red light in intersections, then averages or removes the redundant data according to parking status of the floating car. Experimental results show that this method can deal with the drift data and redundant data in traffic information collection effectively.Secondly, this thesis studies on an improved Newton position-time interpolation(NP-TI) method to sovle the problem that the linear position-time interpolation(P-TI) method can t estimate the floating car s link travel time(LTT) accurately. This method applied the nonlinear Newton interpolation to the P-TI method based on LP-TI to approximate floating car s nonlinear traveling curve in road boundary. Experimental results show that the improved NP-TI method can further improve the effect estimation of the floating car s LTT.Finally, this thesis researches an strategy named AR* guidance strategy and EBKSP guidance strategy to sovle the "congestion drift" phenomenon in traffic guidance strategy. By detecting congestion sections, selecting and dividing the rank of the candidate guidance vehicle, AR* and EBKSP guidance strategies guide vehicles to the optimal path or optimal K paths which have minimal total cost functions of travel time and road counters. Simulation results show that the improved guidance strategies can reduce the phenomenon of "congestion drift" and LATT of the road network.