Wavelet-based electromechanical mode shape identification from ambient data of random load excitation

The wavelet transform method is applied to identify the electromechanical mode shape properties of power system from the ambient data of random load excitation.The relationship between wavelet power spectral density and modal eigenvectors is derived and the wavelet-based coherency function is proposed to discriminate the mode shape of modes with nearly same frequency.The simulation result of a two-area four-generator power system under stationary random excitation demonstrates the effectiveness of the proposed wavelet transform method in modal parameter identification and it also shows that the wavelet method is more effective in identifying modes with weak energy.