With the increasing integration of high-altitude areas and distributed power sources, the issue of PT (Potential Transformer) faults in power systems has become more prominent, particularly the accurate identification and effective suppression of ferroresonance and arc grounding faults, which are critical to ensuring the safe and stable operation of the system. This study proposes a multi-criteria fusion strategy for PT fault identification based on wavelet transform signal processing methods. Firstly, by analyzing the voltage and current signals in the power system, ferroresonance and arc grounding faults are accurately identified, and singular point features are extracted from the signals. Multiscale analysis and wavelet packet decomposition methods are then employed to calculate energy features at different frequencies, constructing a high-precision, multi-criteria fusion fault identification model that enhances both the accuracy and robustness of fault detection. By optimizing the decision thresholds based on actual measured ferroresonance waveforms, the precision and stability of fault identification are further improved. Through simulation and experimental validation, the proposed fault identification and suppression strategy demonstrates good performance in complex power system environments, providing a practical and feasible optimization solution for PT fault prevention and control, ensuring the safety and stability of system operation, and offering solid technical support for the efficient operation and precise measurement of the power system.