A new prediction model of solar radiation based on the neuro-fuzzy model

Abstract

ABSTRACTIn this work, we proposed a new prediction model-based neural network (NN) and neuro-fuzzy systems. The mentioned model is applied over the solar radiation signal which has been used as a new clean energy source recently. The proposed forecasting model consists of two main steps, namely feature selection and forecast engine. In the first step the best candidate inputs are selected and used as input of the next section. The main advantage of the proposed forecasting engine is that there is no need for any past information of the appearances of the input time-series in order to forecast their next values. This model is combined with an intelligent algorithm to improve the abilities of prediction. Efficiency of the proposed method is applied over real-world engineering test cases. The obtained results prove the validity of the proposed model.