Integrated Navigation And Line Regularity Information Based Train Positioning System Design And Implementation

Railways are the backbone of the comprehensive transportation system and important support for developing the modernized economy.In recent years,with the rapid development of communications and artificial intelligence,the level of intelligence and independence of rail transit has been continuously improved.Real-time,fast and accurate acquisition of train running data provides data support for the trial operation test of new lines,and provides a technical foundation for the realization of intelligent and autonomous development of rail transit.The main sources of the existing train operation control system to obtain information,such as train position and speed,are speed sensors,odometers,balise,and track circuits.Considering the unaccessible to train control system during the test,and the demand for continuous high-quality positioning information for future train autonomous operation,this paper proposes the design and implementation of a train positioning system based on the fusion of integrated navigation and line smoothness information.The main contents are as follows:Firstly,as a single sensor cannot continuously and accurately obtain train running data due to the complex train operating environment,a multi-sensor information fusion method based on Extended Kalman Filter(EKF)is studied.Considering the influence of train operating environment noise and the real-time acquisition of train running data,a Global Navigation Satellite System/Inertial Navigation System(GNSS/INS)loosely combined structure information fusion model is established.What is more,an adaptive EKF algorithm is designed for the data fusion.The simulation results show that comparing with a single system the accuracy of train positioning is significantly improved.Secondly,due to the long-term loss of satellite signals,the accumulated error of the INS cannot be effectively modified during the integrated navigation process.Based on the specific fluctuation of the line regularity information,a train regularity based dynamic time warping(DTW)algorithm is proposed.The corresponding relationship between train line regularity and location is revealed.Considering the influence of train positioning accuracy and operating environment noise,an adaptive DTW-Kalman Filter method is proposed to dynamically match and locate the time series of line pitch angle.As a result,the positioning information implements as a virtual balise,which is used to calibrate the INS data and reduce the train positioning error.Finally,according to the needs of practical engineering and the research on train positioning mentioned in the first two parts,a train positioning system that based on the fusion of integrated navigation and the line regularity information is constructed.By completing the work of hardware design,software development,and computerized the train positioning algorithm,this paper implements the integrated management of the acquisition,processing,encrypted transmission,and interaction of train in-transit operation data.And it provides an independent solution for continuous and accurate acquisition of train running information.Furthermore,by designing the multi-source information fusion software interactive interface,the visualized operation of the train positioning system is carried out.This thesis has 54 figures,8 tables and 57 references.