Study On Photovoltaic Output Prediction Considering Meteorological Factors

Photovoltaic power generation has obvious intermittency and volatility,which brings severe challenges to the dispatching management of power system.Photovoltaic power prediction is a kind of economic and effective solution to this problem.Because the output characteristics of photovoltaic power have different periodic fluctuation in different time scales,it is necessary to study the multi-scale prediction method of photovoltaic power.This paper mainly carries on the following aspects of work:Firstly,the influence degree of meteorological factors on photovoltaic output is analyzed qualitatively.Based on the regression analysis and grey correlation analysis,the degree of influence factors of meteorological factors is quantitatively analyzed by historical data.The regression equation of PV power generation power and meteorological factors and the regression equation between meteorological factors are established,and the influence coefficient is obtained according to the regression equation.Secondly,the multi variable grey model modeling method based on the meteorological factors with the influencing factors coefficient is proposed in sunny weather.The weather factors are taken as the influencing factors and the sample solar volt output power as the affected factors variables,and the two are taken as the input variables together,and the photovoltaic output power of the day to be measured as the output is proposed to establish the hourly PV prediction model.At the same time,the prediction accuracy of the model in different seasons is verified.Then,considering that the photovoltaic output will be affected by cloud clusters in cloudy weather,a correction method of solar irradiance based on BP neural network is proposed.The outer atmosphere radiation is selected as the main influencing factor of actual solar radiation receiving.The transmittance and cloud amount of cloud are used as auxiliary factors.The irradiance,the transmittance and cloud amount of the cloud at any time are regarded as BP neural network.The input factor of the solar irradiance is corrected by taking the actual received irradiance as the output factor.Finally,a modeling method combining solar irradiance correction model and multivariable grey model is proposed.The revised solar irradiance is one of the influencing factors.The PV output data of sample day is taken as the affected factor variable,and the photovoltaic output of the day to be measured as the output,and the minute PV output prediction model is established.At the same time,the prediction model is verified in different seasons prediction accuracy.The simulation results show that the model is adaptable to the forecast of sunny and cloudy weather,and the prediction accuracy is in accordance with the standard.