太阳能预报方法及其应用和问题

A Review on Methods of Solar Energy Forecasting and Its Application

太阳能预报包括预测太阳辐射量和光伏发电功率,对光伏发电系统并网运行有重要意义,是当前太阳能开发利用的一个关键问题。本文对国内外太阳能预报方法进行了扼要的评述,归纳了太阳能预报的机理及其方法在光伏发电中的应用。太阳辐射的预报方法主要有传统统计、神经网络、卫星遥感和数值模拟等方法。文中基于光伏发电应用的需求,分析了不同预报方法的优点和不足,并探讨了若干有待进一步改善的问题,展望了国内太阳能预报技术方法的发展和应用前景。

Solar forecasting,consisting of solar radiation forecasting and photovoltaic solar power forecasting,is important for photovoltaic power generation systems in network operation.In recent years,with the development of the solar industry,the demand for solar energy forecasting is increasing.Solar energy prediction methods have been developed in developed country.Our solar photovoltaic technology research is,however,at a primary stage,with only a few universities and institutes conducting simulation-based research,little of which accounts for meteorological factors.According to predicted solar physical factors,the prediction can be generally divided into two categories.One is to predict solar radiation which requires the calculation of photovoltaic power according to the output photoelectric conversion efficiency.The other is direct prediction of output power of PV systems.As the domestic forecast on solar energy technologies and applications are rarely reported,mechanisms of solar forecasting,methods and applications in photovoltaic power generation were reviewed based on the demand for photovoltaic applications.This review would provide an important basis for domestic solar photovoltaic power generation development.This paper focuses on the situations of solar energy prediction at home and abroad,and summarizes the principles of solar energy forecasting and prediction methods.Among them,solar radiation prediction methods involve traditional statistics,neural networks,satellite remote sensing,as well as numerical simulation methods.The photovoltaic power generation forecast method is mainly statistical.Based on the application of photovoltaic power generation needs,different advantages and disadvantages of some forecasting methods were analyzed.Meanwhile,a series of problems in further research on solar forecasting techniques and its domestic development were investigated.The main method is the combination of satellite data,model predictions,meteorological observations data,statistical extrapolation methods,and the neural network.But the numerical weather model seems to be difficult to forecast.The solar forecasting technology includes model results,satellite data,and ground observations