Ultra Short-term Prediction Of Photoelectric Power Considering The Influence Of Fog And Haze

Solar energy is an inexhaustible energy for human beings.Utilizing solar energy to generate electricity is an important way of utilizing solar energy.Accurate photovoltaic power prediction can promote the safe and stable operation of power grid,improve power quality,and provide a basis for grid dispatching,and make response measures to power fluctuations in advance.Due to the influence of meteorological factors and the characteristics of photovoltaic power generation system itself,the photovoltaic power has the characteristics of randomness,fluctuation and intermittence.The integration of photovoltaic power generation will impact the large power grid.In recent years,there have been many long-term and large-scale fog and haze weather in China,especially in winter and spring.Photovoltaic power generation has been severely tested by the frequent fog and haze.At present,there are many studies on the prediction of photovoltaic power in academia,but only a few scholars consider the influence of haze on photovoltaic power.The impact of haze on photovoltaic power generation is determined,but the magnitude of the impact is not yet clear.Therefore,it is of great practical significance and theoretical value to study the correlation between photovoltaic power and various impact factors,especially haze,and to achieve accurate prediction of photovoltaic power generation.The main research work and conclusions of this paper are summarized as follows:(1)Study on the correlation of influencing factors of photovoltaic power generation.By studying the correlation between solar irradiance,photovoltaic plate temperature,ambient temperature,weather quality(AQI)and weather conditions and photovoltaic power,the influencing factors of photovoltaic power are identified.The results show that solar irradiance has the greatest correlation with photovoltaic power,followed by photovoltaic panel temperature,environmental temperature and weather conditions;AQI has a negative correlation with photovoltaic power.(2)The trend analysis of influence factors of photovoltaic power based on Grey correlation.Based on the normalization of 0-1 interval and maximum values of data,the correlation and correlation coefficients of each influencing factor and photovoltaic power are calculated by grey correlation analysis method,and the trend of each influencing factor is analyzed.The results show that,considering the numerical characteristics and stability of the results,it is more appropriate to select the 0-1 interval normalization method when calculating the grey correlation degree and factor weight coefficient of each influence factor by using a single normalization method than the maximum normalization method.The order of the influence factors on photovoltaic power is solar irradiance,AQI,photovoltaic plate temperature and ambient temperature.Degree and weather conditions.(3)Design of photovoltaic power model based on BP neural network.On the basis of data fuzzy processing,abnormal data elimination and normalization processing,a photovoltaic power generation model based on BP neural network is designed.(4)Ultra-short-term prediction analysis of photovoltaic power generation considering haze effect.Based on the operation data of photovoltaic power station in Shenyang Agricultural University,the ultra-short term prediction models of photovoltaic power generation based on traditional BP neural network(considering the influence of haze),BP neural network based on similarity time and BP neural network based on similarity time and haze effect are calculated and analyzed by using the software of matlab.The experimental results verify the validity and feasibility of the model.The prediction accuracy of the three models is 20.83%,14.27% and 12.17%,respectively.The BP neural network prediction model based on similarity timer and haze effect has the highest accuracy.