Research On Detection Methods Of Harmonics And Interharmonics In Power Grid Based On Compressed Sensing

With the advancement of smart power grid construction and development of modern electronic technology in China,an increasing number of nonlinear loads are connected to the power grids,which leads to some power quality problems like the harmonic and inter-harmonic pollution in grids getting worse.In the environment with high-density data,the traditional Nyquist-based sampling method and data compression algorithm have no advantages on harmonic measurement.To monitor the harmonic and inter-harmonic pollution in real time,the sharp increase in the number of terminal devices will bring a huge amount of data,which will consequently put pressure on data transmission and storage.Compressed Sensing theory(CS)provides a fresh idea for data sampling and signal compression,which extremely reduce the demand on the hardware at sampling side and the complexity of compression programs.In this thesis,the features of harmonic and inter-harmonic measurement and the advantage of Compressed Sensing are combined to present a CS-based method for harmonic and inter-harmonic measurement.Application prerequisites and key points of Compressed Sensing are discussed in this thesis.Firstly,the sparsity of the harmonic signal in grids is analyzed to build a basis for the application of this theory and FBD power theory is put forward to realize more increase in the sparsity of harmonic signal.Secondly,according to the specific conditions of harmonic measurement,some existing compressed sensing reconstruction algorithms are analyzed.SLO reconstructing algorithm which has a high reconstructing accuracy is selected to make nonlinear optimization so that it possible to improve the reconstruction accuracy and reduce the running time.Finally,in order to increase the detection accuracy of harmonics and inter-harmonics,window function and double-spectrum-line interpolation algorithm is applied to frequency domain transform under the DFT sparse basis.It solves the problem of spectrum leakage and barrier effect that are likely to appear in the harmonic detection process in the inter-harmonic environment,and greatly improves the detection accuracy of harmonics and inter-harmonics.According to the simulation experiment,the algorithm proposed in this thesis can realize high accurate reconstruction of harmonics and inter-harmonics with a 90%signal compressing rate,which makes the Signal-to-Noise Ratio of reconstructing signal reach 78dB.What's more,the high accurate measurement,which is carried out on the frequencies,amplitudes and phase positions of noisy harmonic and inter-harmonic signal at the compressing rate of 80%,is consistent with the relative requirements listed in the national standards of GB/T14549-1993 and GB/T24337-2009.