Harmonic Analysis In Power System Based On Compressed Sensing

Power quality information collection and data analysis is the basic support to ensure the quality of power supply, improve the efficiency of grid operation and build future smart grid. Collection and analysis of harmonic information is very important among all kinds of power quality. High-density information collection make the amount of data exploding. This research areas attempts to introduce compressed sensing technology into harmonic information collection to overcome the defects of the data compression in conventional Nyquist sampling and to overcome the data storage problems in high-density information collection, which more adapts to the harmonic monitoring environment in power system.After studying, exploring, simulating and analyzing the selection of sparse base, measurement matrix and recovery algorithm in compression acquisition and reconstruction process of harmonic data, this article shows a "harmonic analysis framework based on compressed sensing " and proposes a “compressed sensing improving recovery algorithms for harmonic signal”, with the research status as its foothold.Firstly, on the basis of the analysis that harmonic signal have well sparsity in DFT-based, the paper proposes a harmonic analysis framework through compressed sensing. The framework can achieve the harmonics detection in signal reconstruction process. The experiments show that detection results through harmonic analysis based on compressed sensing proposed in this paper can meet the accuracy requirements of national standards. Specifically, the detection error of frequency, amplitude and phase remains within 10-3,0.15% and 0.1o in this framework when the compression ratio is 30%, which meets the equirements of national standards.Secondly, this paper analyzes the harmonic signal "fundamental component sparsity" theorem by taking harmonic signal features into account in actual power system that the amplitude of the harmonic components is far below the energy of the fundamental wave component. On the basis of this theorem, this paper presents an recovery algorithm for harmonic signal to enhance the detection accuracy of the harmonic components, which achieves the performance optimization for harmonic component by filtering out the fundamental component in compressed signal. Experiments show that, this recovery algorithm can reduce 0.001 Hz in frequency detection error, 0.15 percentage points in amplitude detection error and 0.24 o in phase detection error. At the same time, it enhances the signal reconstruction SNR by 2~8dB compared to the SPG algorithm. The above shows that the algorithm can further improve the performance of detection and reconstruction.