Comparison of fast fourier and wavelet transforms with new improved walsh transform for power components estimation

Abstract

The paper present comparative study of two major algorithms use in measurement of power components in both sinusoidal and non-sinusoidal load conditions i.e. fast Fourier and Wavelet transforms with new improved Walsh function algorithm. The growing use of nonlinear loads in distribution networks lead to disruption of power supply system resulting in distortion of voltage and current waveforms. Measurement using fast Fourier transform (FFT) which is adopted in IEEE standard 1459-2000 is no longer realistic in non-sinusoidal load condition due to its sensitivity to spectral leakage and picket fence phenomenon. Wavelet transform have huge computational demand and also experience spectral leakage effect. The new improved Walsh function takes advantage of simple procedure to develop algorithm to estimate power components. The algorithms are simulated on a model developed on Matlab Simulink. The results were compared with FFT approach and Wavelet packet transform technique and it showed that the algorithm has potential to effectively estimate active, reactive and distortion powers of network under distorted load condition better than FFT. The algorithm is computationally less cumbersome when compared with Wavelet transform. Keywords-sinusoidal and non-sinusoidal waveforms; Walsh function; fast Fourier transform; wavelet transform.