Study of map matching in vehicle tracking system based on the current statistical model

In order to improve precision of vehicle tracking system and reduce the position error of DGPS, authors first analyzed the motive process of vehicle on the road and its moving model. Then, the current statistical model was proposed as moving model of vehicles because in a city environment, vehicles were restricted to travel on its road. Authors proposed a map-aided algorithm of position nearly selected and velocity angularly selected to modify Kalman filtering. The result shows that the accuracy of the whole vehicle tracking system has obviously been increased.