Advanced Modal Identification of Local Electromechanical Transients via the Modified Complex Morlet Wavelet Transform

This work introduces an algorithm for identifying frequency modes in local electromechanical transients. The algorithm is founded on the principles of the Complex Morlet Wavelet Transform, offering an optimized tool for detecting parameters such as frequency, amplitude, phase, and damping constants in non-stationary multimodal signals. To validate the effectiveness of the method, computational experiments were conducted using a synthetic bimodal signal and two real multimodal signals generated from the Power System Stabilizers of different generators at the Tucuru´ı Hydroelectric Power Plant in Brazil. The results demonstrate superior accuracy compared to Deterministic ARMA, PCA, and Prony's method.