Harmonic Detection of Power System Based on SVD and EMD

Abstract

It is greatly significant to detect harmonic accurately and effectively for improving the quality of electric energy in the power system. Actually, much noise exists in signal besides harmonics, inter-harmonics, so the key is how to detect harmonic signals from the complex power system. According to the analysis of harmonic and noise, a new method is proposed in the paper to detect harmonic based on the Empirical Mode Decomposition (EMD) combination with the Singular Value Decomposition (SVD). Applications of three cases into the harmonic wave detecting show that the EMD based on SVD successfully conquers mode aliasing caused by noise. It is effective to detect amplitude and frequency of harmonics from complex signals of power system, proves that the algorithm could effectively and accurately detect harmonic signal.