Output Forecasting Of Photovoltaic System Considering Uncertain Factors

With the development of the national renewable energy, more and more solar technologies have been applied to power system. However, solar energy power is influenced by radiation, which is mostly characterized by short time fluctuations as it responds to various uncertainties of the sky conditions. When integrating this type of source into an existing power distribution system, system planners need an accurate forecasting of the power output, so as to do grid planning and scheduling as well as guarantee the network reliability and stability.The generating capacity of solar energy depends on the radiation and the sky conditions. Solar resources are known to exhibit a high variability in space and time due to the influence of other climatic factors such as cloud cover. And even under the same condition of radiation, the PV output could not be considered as a fixed value, which fluctuates upon temperature, humidity and so on. Therefore, in this paper, uncertain factors were fully considered, and a solar power output forecasting model was given through the forecasting of radiation.Firstly, in this paper, the principle of photovoltaic power generation was analyzed, so as the factors impacting power output; then through analyzing various factors which influence sun radiation, a physical model for calculating radiation was established. Secondly, based on this model, considering the fuzzy division of cloud and its randomness, Markov chain was implied to simulate the cloud sky situation to forecast the radiation. Thirdly, under the cloudy condition, every piece of cloud flows fast, so Fuzzy Markov chain was implied to forecast radiation. After comparing the results and accuracy of the two methods, a conclusion that Fuzzy Markov chain is more capable to forecast cloudy sky radiation was laid. Finally, a Least Squares Fitting model was established to calculate PV output through radiation considering various uncertain factors.