Enhancing wavelet based monitoring

摘要

This paper presents a new technique that can be implemented for enhancing wavelet-based monitoring techniques. A Windowed Wavelet Transform is proposed to increase the accuracy of the wavelet-multi-resolution decomposition and reduce the number of the expansion coefficients that have non-zero magnitude. Furthermore, the maximum wavelet decomposition level assigned for each wavelet can be extended to simultaneously capture high and low frequency contents embedded in any signal. Fast Fourier transform (FFT), short-time Fourier transform (STFT) and Wavelet Transform (WT) are discussed and compared with the proposed technique. The capability of monitoring multi-disturbances that may exist at the same data-window is investigated by implementing the proposed technique in monitoring the interaction between capacitor banks in a distorted distribution system.