A new combined method for RMS calculation based on wavelet packet and Hilbert transform

Abstract

Among several methods used for calculating the root mean square (RMS) of electrical signals, Fourier and wavelet transforms are the most common approaches. The latter also has the advantage of being able to analyze both stationary and nonstationary signals. However, in the wavelet-based methods, the presence of both odd and even harmonics in the input signal causes the harmonic components not to be in the center of the extracted frequency bands and this will reduce the accuracy of the RMS calculation. In order to remove this drawback, this paper proposes a new method based on wavelet and Hilbert transforms, in which the frequency of all harmonic components is increased by half of the main frequency by using a preprocessing technique. In simulation results, the RMS value of a real signal of the steel electric arc furnace of the Esfahan Mobarakeh Steel Company is calculated by using the suggested method. The results clearly show that the accuracy of the proposed approach is better than that of conventional and grouping methods.