Research On Localization Algorithm Of High-speed Mobile Node Under Motion Path Constraint

With the proposal of the major development strategy for a strong transportation country in China,the transportation industry has made significant progress,and the railway and train industry has also achieved remarkable results.It plays a crucial role in promoting the development of China’s modern economic system.However,while developing rapidly,it also faces some problems.With the continuous expansion of railway transportation capacity and the increasing frequency of train departures,the safety of railway operation has gradually become a research hotspot.The key to railway safety is to achieve real-time and accurate positioning of trains,and understanding the real-time position of trains during operation is a prerequisite for ensuring train safety.In recent years,the combination of Global Navigation Satellite System(GNSS)and Strap-down Inertial Navigation System(SINS)has become a commonly used positioning technology for trains.This technology has advantages such as convenience,speed,and low cost.Although satellite navigation positioning has many advantages,there are still many constraints and shortcomings in the practical application of train positioning.During the positioning process,four or more satellites are normally required to complete the positioning.Ground users are easily affected by the surrounding environment when receiving signals from satellites,such as tall buildings,trees,station yards,and tunnels,which can block the satellite signals,causing the required satellite signals to be interrupted and the cumulative errors of SINS cannot be corrected in a timely manner,resulting in significant or even failed positioning errors for trains.Therefore,how to achieve train positioning in a satellite limited environment has become a research focus.To address the above issues,this dissertation proposes a high-speed mobile node localization algorithm based on trajectory constraints.The main research content of the dissertation is as follows:(1)Firstly,the wireless sensor network(WSN)node deployment scheme for railway track lines is designed.And error analysis was conducted on the positioning method based on time difference of arrival(TDOA).At the same time,a motion trajectory constraint model was established,and through simulation experiments,the improvement effect of the motion trajectory constraint assisted positioning model on TDOA positioning accuracy was obtained,with an average improvement of 2-3 meters in positioning accuracy.(2)A TDOA based GNSS/SINS enhanced combination train positioning method,L-TDOA,is proposed to address the problem of low positioning accuracy in obtaining real-time positioning information for high-speed trains on sections with global navigation satellite system failures.Firstly,the deployment structure of WSN along the railway was constructed,and then the trajectory constraint assisted localization model was introduced.Finally,an enhanced combined localization algorithm based on L-TDOA for GNSS/SINS is proposed.Firstly,the initial positioning information of the train is obtained through TDOA,and then the motion trajectory constraint assisted positioning model is used to correct the initial position.At the same time,the continuous positioning information obtained from SINS solution is combined.Finally,Kalman filter(KF)is used for information fusion to achieve accurate positioning of the train.The simulation results show that the proposed algorithm can achieve a positioning error of about 1m in the north direction and about 2m in the east direction,which has a significant effect on improving positioning accuracy and can meet the basic requirements of train positioning.It has certain reference value for achieving high-precision positioning of high-speed trains.(3)Due to the poor observation environment of satellite signals encountered randomly during train operation,it may lead to low accuracy of train positioning results or GNSS positioning failure.An enhanced combination localization algorithm based on trajectory constraints and support vector regression(SVR)is proposed.This method combines the predictive advantages of SVR,while considering that changes in train motion state may lead to the inapplicability of the SVR model.Further enhance the accuracy of train positioning by introducing trajectory constraints.Here,the correction of SINS is used as the training target for SVR.The training samples include the changes in historical SINS and the correction of SINS.By inputting the current SINS variation into the trained model,the corresponding SINS correction can be obtained,greatly improving the positioning accuracy of the train.The simulation results show that the average positioning error of the proposed algorithm in the north direction is around 2.6m,and the average positioning error in the east direction is within1 m,which meets the positioning requirements of the train and provides certain reference value for achieving continuous high-precision positioning of the train.