Research On Train Positioning Method Aided By Track Digital Map

With the development of railway transportation, train density and speed have increased greatly. It makes increasing requirement for train positioning accuracy and reliability. Track digital map has been widely adopted in map-matching due to its low cost and reliable characteristic. However, map-matching doesn't deal with GPS computations and simply obtains the navigation solution from GPS, so it becomes invalid without valid GPS output. Thus, it's significant to use the accurate track digital map to develop a train positioning method with higher performance.The particularity that train moves along fixed trajectory makes it possible to use track map directly in train positioning solution. Firstly, a map-aided GPS integrity monitoring algorithm called MA-RAIM is proposed. A new local track-based coordinate system is applied, and the track geographical information provided by track map database is adopted for redundancy equation and constraint in this algorithm. The erroneous GPS measurement will be eliminated from the computation if it exceeds the threshold limit. This algorithm not only ensures the positioning precision, but also relieves the high requirement for satellite number claimed by conventional GPS integrity monitoring. On the basis of MA-RAIM, this thesis concentrates on an innovative map-aided train positioning approach which integrates track map data, GPS satellite ephemeris and raw measurements in positioning computations. In this key approach, a map-aided position measurement model and a map-aided velocity measurement model are proposed, in which Least squares method and Kalman filter are respectively adopted to calculate the position and velocity of train. The final solution is integrated by assigning suitable weights to the two mentioned models. Detailed mathematical deduce and simulation for the proposed method are given in the thesis. On the basis of simulation, the experimental verification for the proposed method is carried out, and the performance is evaluated and analyzed. The results show that, the train positioning accuracy and stability for the proposed method is obviously improved under the same conditions, compared to the current solution both in normal and signal shadowed environments.