Research On The Key Technique Of Positioning And Orienting System Of Land Vehicle

ALV positioning and orienting system is the foundation to achieve the autonomous navigation functions, the accuracy and reliability of positioning and orienting system influences the entire autonomous driving performance. In order to improve the accuracy and reliability of ALV positioning and orienting system, we have done some research work on DGPS, DR and DGPS/DR integration technologies, all our work is based on vehicle experiment.In the dissertation, we put forward a orienting method based on Differential GPS by analyzing the vehicles kinematical characteristics. The simulation on the real vehicle data showed that this method can offer a very accurate heading result under good differential signal-receiving circumstances.The dissertation deeply studied the sensors of DR System. By using the advanced RBF Neural Network algorithm and taking the heading result of Differential GPS as a standard, we built a calibration method for the initial direction of electronic compass. The experiment shows that the out-put accuracy of the calibrated compass has been improved a lot. By studying the two-wheeled and four-wheeled vehicle kinematical models, the positioning and orienting model based on the ABS sensor is established. After analyzing the error of the model, the factor of the ABS sensor was calibrated. The experiment with the real vehicle proves that this new model can provide good short-term positioning and orienting results. By studying the vehicle dynamics model, we put forward a orienting method based on the front wheel angle sensor and studied the influencing factors of the model accuracy. The simulation with the real vehicle data proves the feasibility and effectiveness of the new positioning and orienting method.In the dissertation, we established a fusion model for the angle movement and a Kalman filtering model for the line movement in DR. By combining the information of DR sensors, we established a Kalman filtering model which combines the information of the angle movement and line movement, and the simulation with the real vehicle data proves that the accuracy of DR has been improved more. By using the federated Kalman filtering, we combined the information of Differential GPS and DR, the simulation with the real vehicle data shows this method can highly improve the accuracy and reliability of ALV positioning and orienting system.