Study On Train Speed Measurement Method Based On Multi-source Information Fusion

With the in-depth study of my country’s next-generation train operation control system,the new mode of vehicle-to-vehicle communication and interconnection puts forward higher requirements for the results of the train speed measurement system.The multi-source information fusion speed measurement method has the characteristics of high reliability,strong redundancy and high precision.It is a key research direction in the field of train speed measurement at home and abroad.At present,the use of a single speed measurement technology for trains has limitations,resulting in low accuracy of train speed measurement.Therefore,after studying and comparing different methods of train speed measurement,this article proposes a multi-sensor speed measurement scheme that does not rely on trackside equipment.Using the combined speed measurement method of Optical Pulse Generator,Doppler radar speed sensor and acceleration transducer.This paper designs a train wheel slip or slide detection and compensation scheme based on the combination of multi-source information and a train speed measurement algorithm based on multi-sensor fusion.The simulation experiment proves the feasibility of the above scheme,improves the accuracy of train speed measurement,and realizes the purpose of autonomous train speed measurement.The main research contents of this article are as follows:1.Establish a train dynamics model,analyze the speed measurement principles and characteristics of wheel axle speed sensors,Doppler radar speed sensor and acceleration transducer,perform error analysis and mathematical modeling of different speed measurement methods,and establish state equations and observation equations,so as to realize the sensor data simulation.2.The Optical Pulse Generator is the main research object,and the working principle and characteristics of the sensor are deeply analyzed.Aiming at the speed measurement error caused by the wheel diameter wear and train wheel slip or slide of the Optical Pulse Generator,a method of wheel diameter wear correction based on the combination of multi-source information and a train wheel slip or slide detection and error compensation method are proposed.The method is verified by simulation,which can effectively detect the train wheel slip or slide and wheel diameter wear of the Optical Pulse Generator,and effectively compensate the sensor’s speed measurement error.3.Taking Optical Pulse Generator,Doppler radar speed sensor,acceleration transducer as the information source,the train combination speed measurement scheme based on the above information source is proposed.The multi-sensor combination structure of direct switching method,weighted average method and optimal estimation method is analyzed and compared,and the optimal estimation method is selected as the research focus.The Kalman filter algorithm in optimal estimation is studied in depth.Based on the Kalman filter algorithm,a distributed architecture is used to propose a federated Kalman filter algorithm suitable for train multi-source information speed measurement,which completes the validity determination and fusion processing process of the measurement information,and finally estimates the optimal solution for the train speed measurement.In the algorithm,the standard Kalman filter is used to design the sub-filter and the main filter module to achieve the purpose of reducing the white noise interference of the sensor and optimally estimating the operating state.The simulation test verifies the effectiveness of the federated Kalman filter algorithm for train multi-source information speed measurement,and improves the anti-interference ability and speed measurement accuracy of train speed measurement and proves the superiority of the algorithm.4.Based on the above research results,a multi-source information speed measurement fusion simulation platform is designed,and the upper computer software of the platform is developed.Integrating the signal simulation and multi-source information processing process on the platform realizes the visualization of the train speed measurement fusion method,the related sensor parameter model and the interference error are controllable,and the operability of the system is enhanced.This simulation platform is of great significance to the analysis,demonstration and research of the algorithms and schemes in this paper.