Detection And Analysis Of Grid Harmonic Based On Wavelet Transform

With the fast development of power electronic technology in recent years, the excessive use of various electronic equipments and non-linear loads has caused increasingly serious pollution problems, affected power quality and endangered the power system’s safety, stability and efficiency, therefore checking the harmonic waves in the grid is extremely important. Simulation researches of measuring harmonic will be made in this paper.Firstly, the article analyses the reason why the waves arise in power systems and its damage and gives a brief introduction to traditional harmonic measurement ways. Then the research and status quo of theory of wavelet research will be introduced, the continuous wavelet transformation, discrete wavelet transformation and wavelet packet transformation wavelet theory will be described, and selection principle of wavelet and wavelet function in the power system harmonic detection will be provided.Limited by the method of its own decomposition algorithm filter in the harmonic detection, there is an aliasing phenomenon in the reconstructed sub-band frequency band. This paper analyzes three existing approaches of anti-aliasing of the wavelet, displays their characteristics and compares results by different anti-aliasing methods. This paper analyzes the problems of the wavelet packet reconstructing filters, adjusts the structure and verifies the detection results after the packet was rearranged.The transient signals will be established and simulated, and these signals in several cases will be simulated by the approach of multi-resolution wavelet analysis in the paper. The wavelet analysis can precisely detect the moment that transient signal occurs. Meanwhile the harmonic signals with integer, and the exponential decay with the harmonic oscillation are simulated, and the simulation experiments not only accurately reflect wavelets’frequency, amplitude, phase size, but also quickly detect the variation of harmonics when time changes. The paper also examines the combination of the wavelet transform and the windowed FFT through harmonics. By comparing the result, the advantages of wavelet transform in the transient harmonic signals are highlighted.In order to solve the accumulated noise from the signal quantification and transmission, the threshold function for reducing signal noise will be adopted. The paper compares the characteristics and disadvantages between hard and soft threshold function of the harmonic signal used for noise reduction. And a relatively better approach is acquired by choosing the function of different parameters of noise simulation to reduce noise.