Research On Selective Harmonic Detection Method Based On Improved SDFT

With the development and use of renewable energy,a large number of non-linear loads and distributed energy sources are connected to the power system,making the existing power grid structure highly electronic,which worsens the problem of harmonic pollution in the power grid and causes nonlinear The load has multiple and scattered characteristics,which makes harmonic control more difficult.accurate harmonic detection is a prerequisite for the treatment of harmonic pollution in power systems,so it is especially critical to study it.Therefore,on the basis of studying the compensation implementation of existing voltage detection active filters,this paper analyzes various mainstream power grid harmonic detection methods.Secondly,combined with the new features of the current voltage detection active filter implementation process of harmonic voltage control,it points out the necessity and significance of using the selective harmonic detection.It analyzes the shortcomings of the existing detection methods to realize the selective harmonic detection method.There are three main aspects as follows:(1)When the multi-synchronous rotating coordinate harmonic detection method based on reactive power theory is used to achieve the selective harmonic extraction,as the specified frequency increases,multiple controllers are prone to coupling,The parameters are difficult to set,and the amount of calculation is large.(2)The harmonic detection algorithm based on Fourier transform is subject to the periodic delay of the algorithm.Also limited by time resolution and frequency resolution,it cannot fully describe the local time-frequency variation characteristics of power grid harmonics;(3)The harmonic detection method based on intelligent algorithms has high operation complexity,low engineering practical value,and operability unrealistic.In view of the above problems,analyze the superiority of the sliding window discrete Fourier transform(SDFT)in selective harmonic detection process,this paper proposes an improved sliding window discrete Fourier that can be used for voltage detection active filters Leaf harmonic extraction algorithm.The improvement is mainly aimed at the sensitivity of the sliding window discrete Fourier transform to frequency fluctuations.The grid harmonic signal model is used in the sliding window discrete Fourier transform equation relationship.By introducing a variable size sampling window,the phase angle changes As a calculation scale,to solve the problem of non-integer sampling caused by frequency fluctuations,the non-integer deviation generated during the calculation process is corrected by the idea of linear interpolation to improve the detection accuracy of the signal.And the use of coordinate transformation to achieve the use of less calculation data to improve the calculation speed,in a stationary coordinate system(α-β coordinate system)to achieve an improved algorithm and accurate reconstruction of harmonic amplitude and phase information.Through the compensation angle and phase compensation,the non-convergence problem caused by the cumulative error of the phase is solved to ensure the detection accuracy of the harmonic signal.Finally,according to the improvement of the sliding window Fourier algorithm based on the adaptability of its frequency fluctuations,by building a DSP-based test platform,compared with the traditional SDFT algorithm,from the detection accuracy,stability,calculation speed and other aspects Based on the comparison results,verify the accuracy,superiority and operability of the harmonic detection algorithm in this paper.