Research on Partial Discharge Signal Detection Based on UHF Sensor Array and Eigenvalue Behavior Analysis

Partial Discharge (PD) in high-voltage electrical equipment can cause insulation deterioration and is one of the causes of major accidents in electrical equipment. Therefore, PD detection of electrical equipment is necessary for electrical safety production operations. In the PD detection of power equipment, since the acquisition system is often collected online, the signal received by the system includes a large number of random noise signals in addition to the PD signal, which makes it important to detect and distinguish the PD signals from the noise signals. Therefore, this paper proposes a method for detecting Ultra High Frequency (UHF) PD signals from random noise based on the eigenvalue behavior analysis of the PD signal covariance matrix and UHF sensor array. The experimental results show that the pure noise average discriminant value is 1.3. When there is PD in space, the mean value of discriminant values is 16.8, significantly higher than the set threshold value of 1.5, which verifies the effectiveness of the proposed method and provides theoretical basis for subsequent denoising processing of PD signals.