earticle

영어

It is an essential factor in automatic navigation systems of ground vehicles, railroad and aviation fields which require high stability to detect the global position of vehicles with credibility and accuracy. This paper proposes an approach to estimate the global position of vehicles moving in the predefined path with a set pattern. Because the information of location received by GPS is included various errors, the errors are minimized using GPS error filter and map matching to obtain correct information. The velocity profile is one of key information for the system model of moving vehicles. We propose an estimation scheme using adaptive Kalman filter where the system model is derived from the moving pattern of vehicles. Since there is high probability that the position obtained from the adaptive Kalman filter is deviated from actual road and/or railway path, we remove the error by using map matching technique. The proposed scheme was experimented in position detection of train moving along predefined section of railway.