Autoregressive Moving Average Model for Forecasting of Solar Radiation

Solar radiation is the energy received from the sun in the form of various rays such as X-rays and UV rays, whereas solar irradiance is the account of radiation of a particular location, recorded with respect to time. Artificial neural network is the human brain inspired system whose functions are based on experiences (set of input values). So, it produces less error and thereby produces more accurate results. Here, in this study, we have recorded solar radiation at Miami, Florida. The hourly irradiance is recorded for 10 days with 24 h data for each day. With such data recording, we can have the exact knowledge of maximum and minimum radiation received at a particular day, month, year, time, and location. The prediction for eleventh day is carried out with the help of ARMA model. The ARMA model makes it easier and simpler to remove trend, seasonality, fluctuations, etc. and hence produces better results that are useful for forecasting solar radiation.