The Research Of Photovoltaic Plant Output Power Prediction

Photovoltaic power generation as a clean form of energy applications having broad application prospects, Increase the proportion of PV is non-fossil energy future long-term development goals, but PV vulnerable to external factors, the output power are volatile, cyclical and intermittent, so when PV grid to improve permeability, affect the stability of the power quality, through accurate prediction of PV output power, can effectively reduce the impact of PV grid on the grid, realization of PV grid-friendly, reduce the "Abandoned light" Phenomenon, thereby to make full use of solar energy and reduce fossil energy use, achieve the purpose of energy restructuring.In this paper, a study of PV power prediction, by collation the research background and research status of PV power prediction, summed up the shortcomings and problems of PV power prediction research. To solve these problems, this paper did the following work.In order to improve the prediction accuracy of PV power output, process analysis of the PV data, and put forward the concept of the ideal output power, through the PVsyst, to simulate the output performance of the solar cell under different conditions. And with examples, to determine the degree of influence of various factors on the PV power output. Average deviation ratio of each generalized type of weathers is determined according to historical PV output data. Then, to compare similar days and adjacent days by example.In order to determine the best model to predict the power output of PV. In the example analysis, the first comparison of the prediction performance of BP neural network and RBF neural network. The results show that BP neural network prediction performance than RBF neural network. On this basis, it is difficult for the similar days sample reflects PV output power versus time, a BP neural network prediction model of PV output is proposed based on correction of gray system model. Adjacent daily power is smoothed with average deviation ratio used to build GM, to obtain daily total power and its judgment interval. Finally, daily accumulative power is corrected with judgment interval as reference to obtain post-correction hourly power. Effectiveness of the method is validated with a practical example. Finally realize the prediction module of PV Monitor and Control System by C++.