GIS discharge fault diagnosis based on wavelet packet singular spectral entropy and WOA-SVM
Sun Chengbin, Xiang Ding, Zhaocheng Yang, Jian Sheng · 2023
To achieve high accuracy diagnosis of GIS discharge faults, this paper proposes a GIS discharge fault diagnosis method based on wavelet packet singular spectral entropy and whale optimization algorithm optimized support vector machine (WOA-SVM). First of all, this study simulates typical discharge faults of GIS through experiments, and collects ultra high frequency signals during discharge. Secondly, the wavelet packet singular spectral entropy of the discharge signal is extracted as the input for GIS discharge fault type discrimination. Then, compare the fault classification performance of three algorithms: SVM with grid search parameters, SVM with PSO parameter optimization, and WOA-SVM. Finally, the accuracy, rate of convergence and fitness curve of the three methods are compared. Experimental results show that the recognition accuracy of the proposed WOA-SVM achieves 95.48%, which is more than 3.8% higher than other comparison algorithms, and its adaptability is better and its rate of convergence is faster.