Research On Train Speed And Positioning Algorithm Based On Multi-sensor Fusion

With the in-depth exploration and research of the next generation train control system in China,new modes such as vehicle-to-vehicle communication and interconnection require higher accuracy,reliability and safety of train speed measurement and positioning systems.In view of the limitations of existing train positioning technology,this thesis proposes a train speed measurement and positioning solution based on multi-sensor fusion.Taking speed sensors and MEMS inertial sensors as research objects,the train slip/slide detection and compensation algorithm based on multi-sensor combination and multi-sensor information fusion algorithm are designed.The simulation experiment proves the feasibility of the scheme,improves the accuracy of speed and distance measurement,and realizes the ability of the train to position independently.The overall research of this thesis mainly focuses on the analysis of speed measurement and positioning technology and multi-sensor fusion design.In the analysis part of speed measurement and positioning technology,the DF16-type speed sensor and MPU6050-type inertial sensor are taken as research objects,and the working principles and characteristics of these sensors are deeply analyzed.Aiming at the problems of wheel diameter wear and slip/slide error of the speed sensor,a wheel diameter wear correction method based on multi-sensor combination and a train slip/slide detection model are proposed,and the validity of the multi-sensor combined train slip/slide detection model is verified by simulation experiments.Testing experiments on the MPU6050 inertial sensor proved the correctness of the algorithms such as initial alignment,attitude solution,and speed and displacement calculations.And some improvement measures are proposed according to the train’s motion characteristics,so that it can stably output information such as attitude,speed and displacement for a long time.In the part of multi-sensor fusion design,a train speed measurement and positioning solution based on multi-sensor fusion is proposed.Using the measured values of the speed sensor and MEMS inertial sensor as the information source,a multi-sensor fusion system was designed using a distributed architecture.The system is mainly composed of a data acquisition layer,a security assurance layer and a fusion processing layer,which completes the process of measurement information collection,data validity judgment and fusion processing in turn,and the optimal solution of the train speed measurement and positioning is finally estimated.The sub-filter and the main filter module are designed by using the standard Kalman filter algorithm,respectively to achieve the effect of reducing the white Gaussian noise interference of the speed sensor and fusing the sensor measurement data to optimally estimate the train running state.The anti-interference ability and the accuracy of the speed and distance measurement of the train speed measurement and positioning systems are improved,which reflects the superiority of the train speed measurement and positioning algorithm based on multi-sensor fusion.Finally,in order to verify the practicability and feasibility of the train speed measurement and positioning system based on multi-sensor fusion,a CRH2-200 simulation model was established based on Simulink software,and a simulation experiment platform was constructed by combining the data from the Huashan North to Lintong East Station on the Zhengxi Line.The experimental results show that the accuracy of speed and distance measurement of the train speed measurement and positioning system based on multi-sensor fusion meets the requirements of the CTCS-3 train control system.Therefore,it shows that the multi-sensor combination-based train slip/slide detection and compensation algorithm and the multi-sensor fusion-based train speed measurement and positioning algorithm have certain application value.