A Position Estimation Method for Urban Rail Transit Locomotives Considering Electrical Measurement Information

The existing section positioning method for urban rail transit (URT) locomotives is constrained by limited techniques and significant errors. In this paper, the correlation between electrical measurements (EMs) and the locomotive positions (LPs) within the traction power supply system (TPSS) is analyzed, and a LP estimation method considering EM information is proposed. The proposed method consists of two steps. First, the multi-coupling relationship between the LPs and EMs within the circuit model of the TPSS is constructed on the basis of circuit theorems, so that the EM-based calculation values of LPs serve as supplementary measurements. Then, the position estimation model based on robust adaptive Kalman filtering (RAKF) is established in conjunction with the locomotive movement process, in which poor-quality data can be identified through weight recognition, and adaptive factors are constructed based on the innovation and state discrepancy value to achieve optimal LP estimation. The results of experiments conducted on Beijing Subway Line 13A indicate that the proposed method can be integrated with existing filtering fusion algorithms and improve positioning accuracy under different measurement conditions.