Wavelet-based intelligent system for monitoring non-stationary disturbances

This paper presents a wavelet-based procedure that will assist in automated detecting, classifying, and measuring of different power system disturbances. Two pattern recognition techniques are used to evaluate the efficiency of the features of the nonstationary signal in the wavelet domain. The paper also presents a new technique that can monitor the variations of the RMS value and any further changes in the nonstationary signal.