This paper presents an improved map-matching algorithm with a Kalman filter whose state space model is different from the conventional algorithms. The main objective of the research is to apply the advanced Kalman filter in an appropriate way to correct GPS errors more effectively. Based on theanalysis of GPS error, the authors redesigned the model of the Kalman filter used in the map-matching algorithm, and then proposed a novel method to update state vector values and parameters of the Kalman filter utilizinghistorical tracks and road map information. The proposed map-matching algorithm handles the biased error and random error of GPS signals in both the road direction and in the direction perpendicular to the road. The authors examine the performance of the algorithm through a sample application.
Abstract