Power Harmonic Detection Method Based On Wavelet Transform

With the application of the nonlinear electronic devices, harmonicspollution caused by these devices is growing seriously. Active powerfilter is an effective method to prevent and manage the harmonics, inwhich harmonics detection is a key technology. Owing to the inherentnonlinearity, randomness, distributing, nonstationarity and complexity ofinfluencing factor on power system harmonics, many conventionaldetection techniques can't acquire accurate harmonic identification.Because of the excellent time-frequency localization characteristics,wavelet transform (WT) is a new research derection for harmonicsdetection.This paper deeply studies harmonics detection based on wavelettransform. In view of aliasing within wavelets, aliasing cause is clarifiedin ways of physical essence and magnitude-frequency characteristics. Inaddition, aliasing phenomenon and aliasing harm are illustrated. On thisbasis, it proposes a new anti-aliasing wavelet based harmonics detectionmethod which be able to eliminate aliasing through simultaneouslyextracted array elements on the odd or even number coordinates, addedtogether the two sequences after made interpolation separately. The newalgorithm gives both attention between signal preservation and aliasingreduction, it overcomes the inherent bug of traditional wavelet filtersrooted from just only concerning perfect reconstruction and totalanti-aliasing.Under the new algorithm, it gives two plans to detect harmonics, andthen use MATLAB for Simulation. It proves that the new algorithm caneffectively eliminate aliasing with high precision and simple operationwhich provides a vigorous mean for uncontrolled harmonics detection inpower system.