Accurate navigation via differential GPS and vehicle local sensors

Accurate positioning of vehicles yields accurate navigation which helps traffic to move more smoothly. The differential global positioning system (DGPS) is one of the most practical navigation tools in a limited area. This paper describes how to combine and/or fuse the DGPS and vehicle sensors to improve position accuracy. The theoretical background for sensor fusion is the use of the Kalman filter. As an example of the proposed sensor fusion, we combine the optical gyro, wheel speed measurements that may include high frequency noises and the DGPS signal that frequently suffers from interference due to various circumstances. Validity of the method was examined by a real automobile in real circumstances.<<ETX>>