Study On Harmonics Analysis & Detection In Power System Based On Wavelet Transform

In recent years, a lot of harmonic current has produced because of the large increased nonlinear electrical equipments. The problem of harmonics in the power system and the safety of equipment is a challenge topic in the world which causes the attentions by a lot of scholars and engineers. The pollution of the electric power system comes from the loss of reactive power and harmonic, so it is obligatory to solve the issue that harmonic and reactive power compensated. The emphases of harmonic is harmonic analysis and harmonic detection. The main basis and final goal of harmonic studies are harmonic analysis, harmonic detection and suppression.First, the paper introduces the concept of harmonic, harmonic causation of production, hazard of harmonic, and then it investigates the methods of harmonic analysis and detection, such as the method of fast Fourier transform, Prony's method, Pisarenko's method, wavelet transform method and reactive power filter, instantaneous reactive power theories, harmonic detection based on adaptive method, harmonic detection based on neural network, and the various methods are compared. The emphases of paper is that according to wavelet packet decomposition algorithms, the power system harmonic current, voltage and electric power RMS derived from wavelet transform coefficients. Improved harmonic analysis method of two-stage method presented based on the algorithms to solve the interference of the fundamental component and the rest of harmonic groups and decrease the spectral leakage. It is verified by the simulation that the proposed method is accurate and proper to analyze harmonics in the electric power system. Follow, it presents a method based on minimum description length principle for selecting an optimal wavelet packet's function to detect harmonic by the wavelet packet algorithms of decomposition and reconstruction. The function can divide the frequency bands linearly, with constant bandwidth to separate each harmonic from the original current signal in order to eliminate harmonic components.And then, it presents a joint method for harmonic detection and suppression based on instantaneous reactive power theories and wavelet packet algorithms of decomposition and reconstruction, and it is verified by the simulation that the proposed method is a advantaged mean for power system detection and suppression. At last, it investigates application of harmonic wavelet in the power harmonic analysis. According to inproved harmonic wavelet, harmonic wavelet algorithms and active harmonic wavelet algorithms derived from it to analyse harmonic. Simulation and compared by FFT algorithms results show that method of harmonic wavelet are precise and feasible.