Research On BDS/INS Combined Train Positioning System

Compared with any conventional single navigation system,the integrated navigation system has higher positioning accuracy and better performance,the integrated navigation system is the main research direction in the navigation system,Especially integrated navigation systems supplemented by inertial navigation systems(INS),with the development of the transportation industry in all countries in the world,the application of integrated navigation systems to railway trains has become the focus of attention in various countries,the integrated navigation system based on multi-sensor information fusion has military significance in all countries.This topic focuses on the data processing algorithms of the navigation system composed of the Beidou satellite navigation system(BDS)and the strapdown inertial navigation system(SINS).First,this text describes the structure and working principle of the strapdown inertial navigation system and the Beidou satellite navigation and positioning system,analyzes the causes of errors in INS and BDS,introduces the BDS and INS combination model briefly,describes the respective shortcomings of SINS and BDS,establishs a position-speed based BDS/INS integrated navigation system,through simulation,the feasibility of BDS/INS integrated navigation positioning system is proved.In view of the failure of the Beidou signal,this text introduces dead reckoning(DR)technology to achieve blind zone positioning,the entire system uses a federated Kalman filter based on radial basis neural network(RBF)optimization for filtering estimation,which effectively suppresses the filter divergence caused by the inaccuracy of the noise statistics of the combined system,For the data fusion algorithm of integrated navigation system,this text proposes a data fusion method based on Radial Basis Function Neural Network(RBF)optimization federated Kalman filter.Radial basis neural network(RBF)is used to implement adaptive information distribution and perform dynamic simulation of data fusion,through the experimental simulation results,it is proved that the optimized federal Kalman filter can improve the positioning accuracy,stability and fault tolerance of the train positioning system.