Research On Harmonic Detection Method Based On Wavelet Transform And Fourier Transform

Electrical energy provided by power system should be a standard sine wave. Amountof harmonic, generated by the nonlinear equipment, has been introduced to the power grid.Among the more and more complicated harmonic components, non-steady harmonic takemore proportion and affect seriously. It is not only harmful to the power system but alsolead harmonic detection to be harder. To ensure stable and secure of System Operation,effective measures must be adopt. The first thing must be done is to detect and analyze theharmonic component exactly.This paper mainly introduces concept about power system, hazard of harmonic andresearch tendency, elaborates the traditional analysis method on harmonic and thedevelopment research on harmonic of world. The analysis focuses on the FourierTransform, Wavelet Transform and Wavelet Entropy theory in the application of harmonicdetection. Taking use of the analysis ability of Wavelet transform in time and frequencydomain, the applicability of Fourier transform in steady harmonic and the sensitivity ofWavelet entropy in dynamic signal, the paper studied a new method which combines threealgorithms.It is to solve the situation that is hard to detect both steady and transient harmonicexactly. Through building signal model and simulating with Fast Fourier Transform,Wavelet Transform and Wavelet Entropy, the paper points out the strengths, weaknessesand scope of application to these three methods and combine the three methods to detectharmonic. The conclusion as follows: For steady harmonic, Fourier Transform is mosteffective method. It has fastest speed, least amount calculation and most accurate result.Wavelet Transform can extract different frequency signals separately and detect unsteadysignals effectively. Wavelet Entropy can determine the moment of disturbance and extractrelevant information.The combined detection algorithm as follows: For the original signal, firstly, the signalis decomposed into the high-frequency part and low-frequency part by the wavelet multi-resolution analysis. The low-frequency part is analyzed by the Fourier Transform toget steady harmonics. And the non-steady harmonic is analyzed by the Wavelet Entropy togain the transition characteristic. This algorithm takes good use of each theory so that it cananalyze both steady and non-steady harmonic well. Through the simulation result to provethat the method can detect Harmonic effectively.