NEURAL-NETWORK-BASED ESTIMATION OF MAXIMUM POWER-GENERATION FROM PV MODULE USING ENVIRONMENTAL INFORMATION

This paper presents an application of artificial neural network for th e estimation of maximum power generation from the PV module. The outpu t power from the PV module depends on the environmental factors such a s irradiation, and cell temperature. For the operation planning of pow er systems, the prediction of the power generation is inevitable for t he PV systems. For this purpose, irradiation, temperature, and wind ve locity are utilized as the input information to the proposed neural ne twork. The output is the predicted maximum power generation under the condition given by those environmental factors. Efficiency of the prop osed estimation scheme is evaluated by using the actual data on daily, monthly, and yearly bases. The proposed method gives highly accurate prediction compared with the prediction by using the conventional mult iple regression model.