New Model for Hourly Solar Radiation Forecasting using for Java Island, Indonesia

Today, the world is looking for sustainable and clean energy, solar systems such as the photovoltaic and solar thermal systems are essential applications for clean energy. The recorded data play an essential role for designing and planning the solar system in terms of saving the cost and enhance the efficiency of the system. But the absence of these data in some places makes it difficult to achieve the optimization in designing and planning for such a system. Thus, in this paper, an Artificial Neural Network (ANN) technique is used to forecast hourly global solar radiation in the Java Island, Indonesia. The recorded data for solar radiation in the past few years are very important to make the prediction in the future. In this paper, an estimation of the hourly average solar radiation in Java Island is developed. This estimation is based on ANN that deals with non-linear data. The actual data of the solar radiation of the location are used as training/testing data while developing the estimation technique.