Study On Voltage Fluctuation Detection Method Of PV System

On the one hand photovoltaic power can make good use of clean energy to ease theenergy crisis and on the other hand, it can be emitted by other local consumption toreduce additional losses caused by long-distance transmission and photovoltaic plant canspare each other with the main grid, which can naturally improve the reliability of powersystem. So photovoltaic power is one of the future direction of energy development, withgood prospects for development. Output of photovoltaic power is influenced greatly bylight intensity, making the utilization of the inverter, which is limiting bottleneck in thedevelopment of photovoltaic power generation.The model of photovoltaic cells is established in the paper and the outputcharacteristics is analyzed. RMS voltage detection method, the fundamental componentdetection method, peak voltage detection method, wavelet analysis and detection method,Hilbert-Huang Transform method and mathematical morphology and the differentialalgorithm detection method are all proved in this paper. Both their advantages anddisadvantages are analyzed. For the end effect of Hilbert-Huang Transform detectionmethod,a new method named mirror method is proposed for data extrapolation,whicheffectively inhibited the end effect, improving the usability Hilbert-Huang Transform.The defect fixed structural elements for mathematical morphology algorithm with thetraditional form of the differential method can not lead to a good deal of noise andproposed the use of generalized morphological opening-closing and morphologicalclosing-opening operation, and in this paper, a new method combined with adaptive themethod proposed a morphological opening and closing operation of adaptive weightedcombination of filter which can design more efficient algorithms to get better results.The effectiveness of the improved methods were tested by numerical example.Voltage sags, voltage temporarily rise, voltage transients and voltage interruptions areanalyzed in order to prove the effectiveness of the improved method. Simulation resultsshow that the improved Hilbert-Huang transform detection method can effectively detectvoltage sags, voltage rise and voltage transients fluctuation but is useless to voltageinterruption. Mirrored data extrapolation method can effectively inhibit the effects ofHilbert-Huang Transform endpoint detection process produced. Improved algorithmbased on morphology and the differential detection method can effectively detect voltage sags, voltage temporarily rise, voltage interruptions and voltage transients, and it canaccurately obtain temporary voltage dips and voltage amplitude rise value. When voltagetransients occurs, it can only be obtained by detecting the instantaneous frequency voltagetransients occurring moments and voltage interruption occurs it can be only obtained bydetect the amplitude versus time curves.