An Adaptive Discharge State Recognition for Optical Partial Discharge Signal Induced by Metal Particles in Gas Insulated Equipment

The assessment of operational risk in gas-insulated equipment can be effectively achieved by identifying the evolution status of defect-induced discharges. However, existing methods are often hindered by electromagnetic noise, which complicates the accurate identification of the intricate discharge features associated with defects. To address this issue, we propose a partial discharge (PD) pattern recognition method based on photon counting features. This approach aims to mitigate the impact of electromagnetic noise and improve the accuracy of defect-induced discharge identification. Furthermore, the discussion of phase-resolved photon counting (PRPC) patterns under the influence of defects is completed, from which we extract feature parameters for quantifying discharge evolution. Subsequently, a clustering strategy to establish a model for recognizing discharge states is proposed. Findings not only contribute to an enhanced understanding of the complex luminescence phenomena related to PD evolution but also offer new perspectives for analyzing defect-induced discharge states.