Forecasting PV power from solar irradiance and temperature using neural networks

Energy is an important aspect in the today's world. Due to the increase in the population and the decrease in oil and other energy resources the power generation using renewable energy has become more popular. Due to the increase in power demand at the power station the forecasting becomes very essential to meet the demand on the daily basis. The amount of power generated depends on solar irradiance, temperature and weather forecast of future. Since there are uncertainties in the forecast it is essential to have very good forecasting techniques to predict the power at the grid stations. In this paper neural network has been used to predict the power generation from the metrological data. In this paper backpropagation algorithm is used for the prediction. The paper presents the forecasting the power and the feasibility analysis of PV system using a grid connected system. In this paper real data from a solar power plant in India is used for analysis. RET screen software has been used for the climatic conditions like humidity, temperature with the radiations. Neural networks model is designed to forecast the power that will be generated for the solar irradiation and the temperature.