Research On Metro Vehicle Positioning And Monitoring Method Based On Acceleration Sensors

Subways and other rapid transit systems are marked symbols of the modern metropolis.The positioning of subway trains is the key to ensuring the safe and efficient operation of trains.Accurately locating and tracking subway trains can help passengers make travel plans and provide operators with auxiliary information about trains.It can also provide convenience for track state prediction,collision early warning system and train abnormal defects and maintenance.In addition,it can also provide a reference elastic calculation data for the positioning scene of some other fixed track features.Therefore,a two-stage automatic real-time positioning metro framework,which uses only consumer-grade low-cost acceleration sensors to realize metro vehicle positioning and monitoring.By deploying consumer-grade acceleration sensors and probes for receiving sensor signals on subway trains and platforms,the cost of complex infrastructure can be saved.In the offline phase,the offline reference map is generated by using the feedback information of the observed acceleration data and the sparse speed displacement between stations.In the online phase,in order to deal with missing data and uncalibrated consumergrade sensors,Gaussian process regression is used to interpolate and predict online acceleration.Then apply the Kalman filter algorithm to combine the acceleration prediction data with the offline reference map to locate and monitor the subway vehicles in real time.We tested the proposed system in Wuhan Metro Line 2,and the results showed that the positioning error of the system was below 5%.By using multiple acceleration sensors for redundant calculations,the problem of cumulative errors and data missing caused by no high-precision industrial-grade sensors is solved.Compared with other deterministic interpolation methods,the interpolation method using Gaussian process regression can not only keep the statistical characteristics of train acceleration as much as possible,but also has better positioning accuracy.The application of the nonlinear Kalman filter algorithm is also more suitable for the construction of the subway model than the linear Kalman filter algorithm.