Research On GNSS/ODO Based Train Positioning Methods

Real-time positioning is an important part of the train control system. An efficient method to improve the positioning accuracy and reliability is using multi-sensor fusion positioning system. GNSS will be introduced to the existing train positioning based on odometer (ODO). It can reduce the error and improve the performance of the positioning system, and achieve the train precise positioning with lower cost. Therefore, research on the GNSS/ODO train positioning methods has important practical significance.The thesis analyzed accuracy and error source of train positioning based on GNSS and ODO respectively, a train positioning system based on GNSS/ODO with embedded processor core is developed. The train positioning information can be collected, synchronized, recorded and processed by the system. An algorithm of wheel slide and slip detection is presented, the algorithm can effectively detect wheel slide and slip by calculating the difference of acceleration measured by GPS and ODO. An algorithm of train wheel diameter automatic calibration based on GPS is presented, the algorithm calibrated the wheel diameter by GPS, and the error of diameter can be controlled under 1%. An algorithm which using current statistical model to fuses the train positioning information by Kalman filter is presented. The accuracy, reliability and availability can be improved effectively by this algorithm.Field test is carried out in Sanjiadian yard of Beijing Railway Bureau by using the train positioning system based on GNSS/ODO. In the test, slide and slip and the changing of diameter are simulated, the algorithm of train positioning information fusion was validated and evaluated in the aspect of accuracy, reliability and availability.