Research On Train Interated Location Technology Based On GSM-R Network

In recent years, with the rapid development of control technology, communication technology, computer technology and railway signaling technology, train control technology is constantly being developed, improved and advanced. Currently, in the CTCS-2and CTCS-3system, a variety of train positioning technologies are applied. Typical train positioning techniques include GPS (Global Positioning System), DR (Dead Reckoning). GIS (Geographic Information System) and RBC (Radio Block Center) and the integrated positioning and navigation system based on them. These techniques have their own features and application fields, and defects as well. Meanwhile, stimulated by the rapid development of cellular mobile communication technology, the mobile terminal location technology based on fingerprint location has been widely used and popularized in2G GSM (Global System for Mobile Communications) network because of its unique advantages.GSM-R (GSM for Railway), a digital mobile communication system evolved from GSM cellular system, is currently used in CTCS (Chinese Train Control System) communication platform. With the assistant of this system, the scheduling and safety management of the high-speed train are guaranteed by the railway control center. Consequently, based on the analysis of GPS. DR. WKNN (Weighted K-Nearest Neighbor) positioning techniques and features of CTCS. a train fingerprint positioning system based on the GSM-R network is designed. In addition, a kind of GPS/DR/WKNN integrated positioning algorithm is proposed, which is outperformed the traditional GPS/DR integrated positioning algorithm. The main contents of this thesis are summarized as follows:First of all, on the basis of reading and analyzing a lot of literature, the functional structure of CTCS-3is understood and mastered, and the merits and demerits of the existing train positioning techniques are also analyzed and summarized.Secondly, the positioning principle and error model of GPS, DR, Query-Balise and the track circuit positioning system are analyzed. Combined with the Kalman filtering techniques, the algorithm of GPS/DR integrated navigation with Federated Kalman filtering is simulated and analyzed. Then based on the GSM-R network characteristics and train positioning requirements, the fingerprint positioning system based on the GSM-R network is designed.Thirdly, with the analysis of typical fingerprint positioning algorithms and GSM-R wireless channel models, the fingerprint positioning techniques based on the GSM-R network is researched.Moreover, on the MATLAB platform, integrated positioning techniques of WKNN fingerprint location combined with track circuits is simulated and analyzed.Finally, by comparing the estimated error covariance to dynamically allocate the coefficients of the subsystem, an adaptive Federated Kalman filter is realized. The DR/WKNN and GPS/DR/WKNN Federal Kalman filters based on the GSM-R network are designed, and better estimation accuracy is obtained by GPS/DR/WKNN algorithm. In the case of GPS working and failure, GPS, DR, WKNN and their integrated positioning techniques are modeled and simulated as well;the excellent performance of the GPS/DR/WKNN integrated positioning technique is evaluated and proved.