Estimating Efficiency of Monocrystalline and Polycrystalline Photovoltaic Panels Using Neural Network Models

The energy production analysis of a  photovoltaic system depends on the panels tempreture and solar radiation. An endless and free source of solar energy received at the Earth's surface depends on the geographical location, different hours of day and seasons of the year.Hence, its correct evaluation is a strategic factor for the feasibility of a solar system. in this paper, a new method of energy modeling of photovoltaic systems is proposed by using the radiation and temperature data obtained from monitoring of monocrystalline and polycrystalline solar panels installed at the solar site of the Vali-e Asr university of Rafsanjan. The model is derived using data in a period of one year of the solar site by ANN models which is trained and tested by a multi - layer Perceptron neural network. The inputs of the model include the temperature of the panel and the direct solar radiation and Its output is the production power of both monocrystalline and polycrystalline solar panels of this solar site. The results showed that, it is proper to chose The activation function at the hidden layers of logsig, tansig, tansig with the number of [10 10 10] neurons.