Abstract: Aiming at conquering the spectral aliasing in the Mallat algorithm, a new method for fundamental wave detection with the wavelet transform and empirical mode decomposition (EMD) is proposed. The discrete dyadic wavelet transform decomposes the harmonic signal into sub-band signals of different frequency-bands, and the optimal decomposition level is determined. Then the single band reconstruction is performed to the sub-band signal including fundamental frequency component, and the fundamental wave can be extracted by empirical mode decomposition. Finally, the fundamental frequency and amplitude of the signal are estimated by the least square method in the time domain. Through the comparison of simulation results generated by different methods and the application, it is shown that the fundamental wave can be extracted effectively with this method, so that the frequency measurement and amplitude measurement are of high accuracy.

Key words: Empirical mode decomposition, Fundamental wave detection, Mallat algorithm, Spectral aliasing, Wavelet transform