Active Power Filter Analysis And Simulation Based On Wavelet Theory

In view of the current power system harmonic analysis methods can't detect the dynamics of harmonic in real-time quickly and accurately . The FFT on the steady-state harmonic analysis and harmonic fluctuations, and the WT Quick detection of changes in harmonic are integrated. Both advantages are highly complementary characteristics, a Fourier analysis will be combined with the wavelet transform, for a variety of harmonic analysis.At present, the main harmonic detective method are those transform domain analysis, they are effective methods of analysis of the signal spectrum, including the Fourier transform and the wavelet transform. The Fourier Transform is built on the basis of the cyclical sequence, so it play an important role on the analysis of steady-state harmonic signals.But what the FT obtained is the average value by the frequency of information throughout the signal time-domain and thus it can not give local time frequency information. In recent years ,the development of the wavelet transform is a time-frequency domain based on the partial conversion, and high-precision band gradually take the time-frequency step size, which can be focused on the analysis of the signal, any details of the mutation can effectively extract information. It is suitable for harmonic Voltage amplitude or frequency over time the impact of changes in the form of mutation analysis. However, due to common discrete wavelet transform binary frequency division of the rough, as well as the transformation can not be directly spectrum of information, it is not suitable for harmonic analysis with more steady signal [1].In this paper, the advantages of Fourier transform and wavelet transform are integrated, put forward a new method which combine Fourier and wavelet transform harmonic detection together. This new method can detect accurate steady-state harmonic information and non-steady-state harmonic component.The harmonic detection algorithm of Fourier and wavelet combining, is to be found for the first signal DWT, the original signal divided into high-frequency (details) and some of the low-frequency (picture). Non-steady-state harmonic as well as mutations, such as the singular point of intermittent signal component in the general wavelet transform some of the details of the future, some of the signals continue to wavelet analysis ; for the steady-state low-frequency part of the picture using Fourier transform analysis, the steady-state out Harmonic component. This can be a variety of transient wave components, can draw accurate steady-state harmonics. In this paper, the use of the FFT and Wavelet Transform combining the harmonic detection method simulation of the system through a simulation of wavelet analysis with time-domain and frequency domain of the double-resolution, to better solve the problem which the Fourier analysis can not. In harmonic analysis, using wavelet analysis method can not only get the right harmonics, but also can solve the transient signal which the Fourier analysis about the weight of the can't extract. Transient component of the time positioning, Voltage, current wave of intermittent, processes, depression and weight of the transient detection have good results.In this paper, we build a new structure of the filter banks, which can put each node and the corresponding band of decomposition in wavelet packages together, and find the arranged law of the filter bank structure, this law can build on many occasions Decomposition of the filter bank structure.