Kalman Based Data Fusion Algorithm for High Accuracy Train Positioning

The track circuit can automatically and continuously detect the train position occupancy in the track section. Most of the existing train positioning systems are limited to determine which block is occupied, but can not obtain the accurate position of the train in the block section. Global Navigation Satellite System (GNSS) has the advantages of high precision and real-time positioning, and it can provide an absolute position data. However, in dense areas of tracks, such as stations, it is impossible to accurately determine where the train is located due to GNSS positioning errors. In this paper, an algorithm is proposed to combine GNSS positioning technology with the track circuit positioning method to obtain continuous train position information in the block areas. The position information of the track circuit can also correct the error of GNSS positioning data, and effectively improve train positioning accuracy. What is more, a train positioning simulation platform is developed and applied to verify the proposed method.