Power System Data Acquisition And Reconstruction Algorithm Based On Compressed Sensing

The current methods of signals' detection and compression in the power systemare based on Shannon sampling theorem. Taking up a lot of disk space, thecollection of enormous amount of data is a great burden to the power system datatransfer and storage. Moreover, most of the information contained in the signal isnot necessary, which makes a big waste of data collection. However, the Shannonsampling theorem to reconstruct the original signal is a sufficient condition, not anecessary one. In recent years, a new data acquisition theory called compressedsensing has been proposed, and it is able to break through the restriction of theShannon sampling theorem for sparse signal. Compressive sensing theory abandonstraditional mode, what is processing after sampling, sampling and compressing atthe same time, recovers original signal through a small number of measurements.This paper studies compressed sensing theory theoretical basis, and designed a dataacquisition and reconstruction system based on the theory of compressed sensing,which is achievable for hardware and particular for the grid data signals, the maincontents of this paper are:(1)Research the foundation theory. Using simulation proved that the grid signalcan be acquired and reconstructed with compressed sensing theory.(2)Research the way in which random demodulator works. According to thecharacteristics of the grid data, built single grid AIC(Analog to InformationConverter) acquisition system framework, with the algorithm and simulationproving the correctness of the system.(3)Research the greedy algorithm, containing the orthogonal matching pursuitalgorithm, regularized orthogonal matching pursuit algorithm, compressivesampling matching pursuit algorithm, and complete the algorithm. The simulationproves that the algorithm can accurately reconstruct the grid signal.(4)Combine the AIC systems and signal reconstruction algorithm into an overalldesign. For a variety of types of grid data, this article used compressed sensingmethod for signal acquisition and reconstruction. Analyzed the simulation resultsand proved that the new system's sample rate is much lower than the conventionalmethod, and can correctly reconstruct the signal.