Data-Driven Identification of Rail Impedance Parameters in Urban Rail Transit Systems

Abstract

Accurate rail impedance parameters are essential for reliable stray current simulation in urban rail transit systems. In practice, these parameters are often determined using empirical values, which may deviate from actual operating conditions. This paper proposes a data-driven method for identifying rail impedance parameters using measured operational data. A return current model of the DC traction power supply system is established, and a rail impedance identification model is developed based on circuit analysis. The measured data are preprocessed using the median absolute deviation (MAD) method and wavelet denoising to reduce noise and outliers. The proposed method is validated using operational data from a Beijing metro traction power supply section. The results show that the rail impedance can be stably identified with an estimated value of approximately 0.02814 Ω/km and an error within 5%.