A novel optimized in Harmonic detection using Reconstructed WPC considering spectral leakage and edge effects

Abstract- Harmonics is an important issue in electric power systems and power quality. However, real-time techniques are receiving increased attention to estimate and quantify harmonics. A wavelet packet transform (WPT) method is introduced as a powerful tool for classification and quantification detection of power quality events. The proposed method, computes the distribution of RMS quantities with respect to individual frequency bands directly from the wavelet transform coefficients. Uniform frequency bands result from the Wavelet Packet Transform decomposition of power system waveforms may be used for identification of harmonic frequency bands. This paper presents a novel harmonic detection and measurement method. The novel proposed method uses combination of two adjacent bands that present highest effects caused by each harmonic. This paper also offers the use of a compensation method in Reconstructed Wavelet Packet Coefficient to prevent energy reduction. In addition, it introduces a modified method to reduce edge effects and spectral leakage. At last, linear optimization is used to reduce error. Methods are simulated and experimented for a real signal in presence of harmonics. Comparative study shows satisfactory results.