Research on Positioning of Permanent Magnet Maglev Trains Based on Weighted Adaptive Kalman Information Fusion

In this paper, an improved adaptive Kalman fusion method is proposed to solve the problems of low accuracy, poor interference immunity and stability when using a single sensor to localize a suspended permanent magnetic levitation train in a practical situation. The measurement data from three sensors are fused in the first layer by weighted adaptive Kalman filtering (WKF). Then, the measurements and the positioning data from the Global Navigation Satellite System (GNSS) are fused in the second layer using recursive weighted least squares (RW) to achieve highly accurate positioning of the levitated permanent magnet maglev train. The experimental results of WKF-RW algorithm improve the positioning accuracy by 51.79% and 32.5% compared with the conventional Kalman (KF) and WKF, respectively, which effectively improves the positioning accuracy and robustness of the permanent magnetically levitated trains and can be better applied to the real environment.