Joint Time-Frequency Analysis of Partial Discharge AE Signals for Pattern Recognition

Abstract

Joint Time-Frequency Analysis is an essential tool in signal processing, especially for developing pattern recognition techniques for condition monitoring and diagnostics. Several available methods in the literature demonstrate the successful application of Joint Time-Frequency Analysis (JTFA) for nonstationary signal processing, partial discharge analysis, condition monitoring, and biomedical engineering. Partial discharge detection and analysis is an essential topic for the condition monitoring of Transformer, Generator, High Voltage Cables, and Gas Insulated substations. There are several tools available for JTFA using both MATLAB and LabVIEW. However, they are limited to the following techniques Short-Time Fourier Transform, Wigner-Ville Transformation, and Wavelet Transform, whose performance reduces under the noise. Hence, there is a need to develop a suitable intelligent tool for real-world applications with superior performance under noise. In this paper, a Joint Time-Frequency Analysis tool has been developed as a first step for the pattern recognition of partial discharge signatures. The developed MATLAB GUI uses an advanced multiresolution analysis algorithm such as Complex S-Transform (CST) for analyzing partial discharge signals. The developed tool has been successfully applied to analyze partial discharges and provides a better result than existing techniques.