Study On The Short Term Prediction Of Photovoltaic Power Based On Combination Model

With the development of science and technology and the progress of human society, the demand of renewable energy is increasing day by day. At the same time, the photovoltaic power generation got rapid development; Photovoltaic power generation, however, influenced by temperature, light intensity, wind speed, geographical position and many factors, so its power output has a great deal of stochastic volatility and unpredictability; Such features in photovoltaic power generation systems will cause certain impact to the grid when incorporated into the power grid; and made it unfavorable to electric power dispatching department to make arrangements. So the prediction of the power on photovoltaic power generation system in advance is very important.This thesis mainly studied the photovoltaic(pv) power prediction technology research status and the existing problems. For single model applied to power prediction problem of prediction accuracy is hard to meet the requirements putting forward the exponential smoothing method and markov chain into the grey model. Respectively established the grey forecasting model based on exponential smoothing and the grey- statistical combination forecast model based on markov chain. The two combination forecast model should be in the actual output power of photovoltaic power generation system prediction and obtained the ideal prediction results. For empirical method is too subjective in the process of the exponential smoothing model, using error test algorithm to the optimal choice of smoothing coefficient, this will be beneficial to the improvement of the model prediction accuracy.In this thesis, the research content including power generation principle and the classification of photovoltaic power generation systems and the influence factors of photovoltaic power output;Photovoltaic power short-term prediction technologies are studied and discussed; General prediction methods of irradiation intensity is studied in degrees. And analysis of exponential smoothing is introduced emphatically, gray model and markov chain model of modeling principle;, throughexamples in this thesis, the combination of the proposed model is used for photovoltaic power prediction, the prediction accuracy compared with the single model prediction accuracy has improved further, it has certain feasibility and reliability, and the proposed combination forecast model can avoid complex irradiation prediction and analysis of meteorological factors, has the characteristics of easy.