Transformer winding short-circuit fault diagnosis based on transient vibration signals

Abstract

To address the limitations of traditional frequency response analysis and short-circuit impedance methods in diagnosing transformer winding deformation, this paper establishes a transformer transient vibration testing platform based on short-circuit impulse excitation. The variations in the time-domain and frequency-domain waveforms, as well as in the characteristic parameters, namely the root-mean-square (RMS) deviation rate and dominant-frequency deviation rate, were investigated under different short-circuit severities. The results show that repeated tests under normal conditions yield highly consistent results, demonstrating good repeatability and stability of the method. Compared with the normal state, vibration-signal amplitudes under short-circuit faults increase significantly, and the dominant frequency shifts toward higher frequencies. As fault severity increases, the RMS deviation rate rises from 9.32% to 14.91%, and the dominant-frequency deviation rate increases from 12.9% to 19.4%. Therefore, this method can be effectively applied to the diagnosis of transformer winding short-circuit faults.