Development and Application of an Enhanced Kalman Filter and Global Positioning System Error-Correction Approach for Improved Map-Matching

Map-matching, which reconciles a vehicle’s location with the underlying road map, is a fundamental function of a land vehicle navigation system. This paper presents an improved Kalman filter approach whose state space model is different from the conventional ones. The main objective of the research is to develop and apply a proper Kalman filter-based model for effectively correcting the Global Positioning System errors in map-matching. Based on the in-depth investigation of the characteristics of the Global Positioning System errors, the authors presents a novel approach to update the state vector and other related parameters of the Kalman filter using both the historical tracks and the road map information. The performance of the proposed approach is thoroughly examined by sample applications with real field data. The result shows that it handles the biased error and the random error of the Global Positioning System signals reasonably well in both the along-road and cross-road directions.