Map Matching Algorithm And Data Transmission Frequency Optimization Of Probe Vehicle System

Probe vehicle technology has been utilized as a new way to monitor road traffic condition by collecting road traffic information using the vehicles equipped with GPS (Global Positioning System) and communication device. There are widely concerns in researches and applications of probe vehicle technology in the world. Map Matching, an important part of probe vehicle technology, is the primary procedure to generate road traffic information from time series GPS data, and has a direct impact on traffic information accuracy. Low transmission frequency data collected by dispatching management system of taxi is used by the current probe vehicle system. It has important theoretical value to build algorithm based on the low transmission frequency data to ensure the accuracy. Usually, the data transmission frequency, road network characteristics and traffic condition have effects on map-matching accuracy, and as the main influencing factors of the map matching accuracy, the data transmission frequency also determines the communication cost. Therefore, the study of data transmission frequency optimization based on the precision requirement of map-matching, is greatly significant to reduce the communication cost and improve the operation efficiency of the probe vehicle system.First, the paper reviews the exiting methods of map-matching and the main factors that can influence the map-matching efficiency. After taking into low transmission frequency data characteristics into account and selecting the candidate links considering the GPS errors for each GPS point, the paper utilizes the shortest path algorithm to select the candidate links from the road network and then determines the final running trajectory for probe vehicle based on the fuzzy logic inference. The proposed map-matching algorithm based on low transmission frequency data is evaluated with extensive field experiment data collected in part urban areas of Beijing city. Second, based on the proposed map-matching algorithm, the map-matching accuracy is calculated with the consideration of main influence factors, such as data transmission frequency, road category, and traffic flow smoothness. Moreover, a map-matching accuracy evaluation model is established using road network characteristics and data transmission frequency as explanatory variables, and the model has a high transferability by considering urban road network characteristics. Finally, based on the established map-matching accuracy evaluation model, a probe vehicle data transmission frequency optimization model is proposed, in which the main constraints such as the expected accuracy, ideal data transmission frequency, and the requirement of accuracy improvement rate are taken into consideration. The probe vehicle data transmission frequency optimization model is applied in some case studies. The result shows that in the studying area, when the expected accuracy of map-matching is larger than85%, and the rate of improved accuracy is more than0.2%per second, the best data transmission frequency is68seconds based on that the data transmission frequency can be achieved1second. The case shows that the model avoids successfully the situation in which the probe vehicle data transmission frequency is lowered too small while the map-matching accuracy is improved a little and indicates effectiveness and certain reference value to probe vehicle system configuration optimization.