Photovoltaic system power forecasting based on combined grey model and BP neural network

With the emergence of energy crisis and environmental pollution, the large scale photovoltaic power systems have been widely applied. However, the output power of photovoltaic power system has the property of uncertainties. In order to lighten the adverse influence for power grid, this paper attempts a method based on Grey combination model to forecast the short-term power output of a PV power system. The proposed method is a combination of grey model and BP neural network model. It takes the main factors of power output of photovoltaic power system into consideration and builds GM(1, 1) model by choosing proper samples, and then builds the BP Neutral Network model using residual error series between fitted values and real values, finally modifies the GM(1, 1) value. The result of test example shows that the Grey combination model can efficiently predict the short-term power output for photovoltaic system and has a potential value in practical applications.