Abstract:Feature extraction and classifier optimization are important parts in the diagnosis of partial discharge (PD). Traditional feature extraction methods generally target not the images but structured data. Traditional classifier optimization methods have many disadvantages such as the many parameters that need to be adjusted, difficulty in determining the best parameters, and the ease in falling into local optimal solutions. To eliminate these defects, a PD diagnosis method of gas-insulated switchgear (GIS) based on the Zernike moment and improved support vector machine (SVM) is proposed. Firstly, five typical PD models of corona discharge, metal particle discharge, floating electrode discharge, void discharge, and surface discharge were designed. Then, the UHF sensor was used to collect discharge signals of different defects and different types of interference signals to build a PD phase-resolved pulse sequence (PRPS) patterns library. After that, the Zernike moment of the PRPS images were extracted as the PD features. Finally, the SVM model based on the differential evolution adaptive bacterial foraging optimization (DEABFO) algorithm was used to fulfill the GIS PD diagnosis. The results show that the Zernike moment feature extraction method is suitable for PD PRPS images, and DEABFO improves the convergence speed of the SVM parameter optimization process and can effectively avoid the probability of falling into the local optimum. Compared with other image feature extraction methods and other SVM optimization algorithms, this method has high accuracy in GIS PD fault diagnosis, with the recognition rate reaching 91.23%.

SVM-based Partial Discharge Pattern Classification for GIS

Partial discharge diagnosis by simultaneous observation of discharge pulses and vibration signal

Partial Discharge Source Localization in GIS Based on Image Edge Detection and Support Vector Machine

Discrimination of Three or More Partial Discharge Sources by Multi-Step Clustering of Cumulative Energy Features

Partial Discharge Pattern Recognition Based on an Ensembled Simple Convolutional Neural Network and a Quadratic Support Vector Machine

Partial Discharge Fault Diagnosis Based on Zernike Moment and Improved Bacterial Foraging Optimization Algorithm

Partial Discharge Recognition with a Multi-Resolution Convolutional Neural Network

Statistical Feature Extraction Combined with Generalized Discriminant Component Analysis Driven SVM for Fault Diagnosis of HVDC GIS

Multi-source partial discharge pattern recognition algorithm based on DCGAN-Yolov5

Application of SCNGO-VMD-SVM in Identification of Gas Insulated Switchgear Partial Discharge

A Pattern Recognition Method for Partial Discharge Detection on Insulated Overhead Conductors

Partial Discharge Diagnostics: Data Cleaning and Feature Extraction

Towards Generalizable Classification of Partial Discharges in Gas-Insulated HVDC Systems using Neural Networks: Protrusions and Particles

Separation and Classification of Concurrent Partial Discharge Signals Using Statistical-Based Feature Analysis

Partial Discharge Inversion Localization Method for GIS Based on Twin Database

Identification of Partial Discharge Sources by Feature Extraction from a Signal Conditioning System

Pattern Recognition of Partial Discharge Faults in Switchgear Using a Back Propagation Neural Network Optimized by an Improved Mantis Search Algorithm

Diagnosis of Partial Discharge in GIS Based on UHF Sensing Technique

Fundamentals in computer aided PD processing, PD pattern recognition and automated diagnosis in GIS

A Novel Partial Discharge Localization Method for GIL Based on the 3D Optical Signal Irradiance Fingerprint and Bagging‐kelm

Research on the Typical Partial Discharge Using the UHF Detection Method for GIS