Research On EKF Based GPS/ODO Integrated Train Positioning Method

The train location subsystem plays a key role in train control system. In the central and western regions of China, railway lines with low density are in urgent need of a safe, reliable and lowcost train control system. The train positioning method based on GPS and odometer(ODO) can effectively save the trackside equipment construction costs as well as railway maintenance costs.This thesis puts forward the EKF-based, GPS/ODO-integrated train positioning method combining the existing GPS positioning method with ODO positioning method. This method aims at reducing the positioning error and improving the reliability of train positioning system. The main contents are as follows.This thesis analyses the ODO positioning velocity error and establishes error model. Moreover, the thesis puts forward a method of slide and slip detection which breaks the limitations of the fixed threshold method applying the small probability principle.The thesis proposes a method of wheel diameter correction combining GPS positioning. With the growth of the train running mileage, the wheel diameter loss continues to accumulate. The appropriate threshold of wheel diameter correction selected in accordance with the wheel diameter loss is more accurate than the traditional method with code dialing device.Under normal running condition of the train, the velocity difference between GPS velocity and ODO velocity is a zero mean random white noise. The thesis establishes the prediction model of velocity difference between GPS velocity and ODO velocity applying time series analysis method. It also attempts to correct the error of ODO velocity during the period of slide and slip.In order to solve the positioning problem during GPS outages, the thesis proposes the train positioning method based on ODO and map data base. This method solves the problem of train positioning during GPS outages when the train is passing through bridges, tunnels and so on.Moreover, the system composition and algorithm design of GPS/ODO integrated train positioning method based on EKF are presented in the thesis. Finally, the method and theory proposed in this thesis are verified by test data of Jincheng Coal Mining Group railway line. The experimental results show that the methods could correct the velocity error of wheel diameter loss and wheel slide and slip. It could also realize accurate train positioning during GPS outages. Combined with multi-sensor information fusion, extended Kalman filter could improve the accuracy of train positioning system.