Research on train positioning calibration technology based on multi-sensor fusion

In order to address the difficulties of satellite positioning for trains in tunnels and mountainous areas, this paper adopts an alternative solution that combines electronic map information with multi-sensor data. The paper analyzes the error mechanisms of different sensors, determines the threshold values for error correction of each sensor, and designs a scheme for error correction and information fusion among multiple sensors. An adaptive Kalman filtering algorithm is used to estimate noise in real-time. This paper further integrates sensor information with electronic map information, using accelerometers to detect slope change points and combining this data with the line information from electronic maps to further refine the train’s positioning. Experimental results demonstrate that this solution converges the positioning scope of the train and enhances the autonomy of train positioning.