Research Of Thetrain Positioning Method Based On Beidou Satellite Navigation System

With the rapid development of high-speed railway,train positioning method based on satellite navigation system with its unique advantages,represented development direction for future technology in the train positioning field.As the core of the train operation control system,train positioning method must be respond to the trend of The Times and go along the development way of safe,efficient and innovative.At present,there are a lot of methods for train poaitioning,but as the change of application environment,exposed the defects of different level gradually.In order to enhance the reliablity of train operatin control system,this paper from two aspects of Beidou satellite navigation system effectively and failure,study the different method of train positioning furtherly.In the presence of Beidou satellite navigation system effectively,this paper proposes the train positioning method based on cubic B-spline curve fitting.Firstly,deal with the measurement data of Beidou satellite and to coordinate transformation.Secondly,using the cubic B-spline curve fit the motion of the train route.Thirdly,using variable step size compound Simpson method calculate the operation of the train position.Lastly,compare cubic B-spline curve fitting train positioning method with the least squares curve fitting train positioning method.The results show that cubic B-spline curve fitting train positioning method accuracy is higher than the least squares curve fitting train positioning method,residual range become smaller and flexibility is stronger.In the presence of Beidou satellite navigation system failure,this paper proposes the fuzzy neural network train positioning method based on GSM-R.Firstly,analyses the coverage intensity and tansmission loss of GSM-R network,collect and calculate related data.Secondly,according to the learing principle of Back-Propagation neural network and fuzzy control,design fuzzy neural network controller in the sysrem.Lastly,by the Simulink simulation experiments,contrast the fuzzy neural network train positioning method based on GSM-R and the Back-Propagation neural network train positioning method.The results show that fuzzy neural network positioning method based on GSM-R accuracy is higher than Back-Propagation neural train positioning method.Convergence speed become faster,east error and north error are in the allowable range of Beidou satellite navigation system positioning error.It can meet train positioning accuracy of high-speed trains.Ensure the real-time and continuity of train positioning in the satellite navigation blind area.