Solar Irradiance Forecasting Using Deep Neural Networks

Abstract

Predicting solar irradiance has been an important topic in renewable energy generation. Prediction improves the planning and operation of photovoltaic systems and yields many economic advantages for electric utilities. The irradiance can be predicted using statistical methods such as artificial neural networks (ANN), support vector machines (SVM), or autoregressive moving average (ARMA). However, they either lack accuracy because they cannot capture long-term dependency or cannot be used with big data because of the scalability. This paper presents a method to predict the solar irradiance using deep neural networks. Deep recurrent neural networks (DRNNs) add complexity to the model without specifying what form the variation should take and allow the extraction of high-level features. The DRNN is used to predict the irradiance. The data utilized in this study is real data obtained from natural resources in Canada. The simulation of this method will be compared to several common methods such as support vector regression and feedforward neural networks (FNN). The results show that deep learning neural networks can outperform all other methods, as the performance tests indicate.