Abstract
This work introduces a novel multi-branch attentive gated recurrent residual network (ResAttGRU) consisting of multiple branches of residual networks (ResNet), gated recurrent units (GRUs), and the attention mechanism. The proposed multi-branch ResAttGRU is capable of modeling data at various resolutions, extracting hierarchical features, and capturing short- and long-term dependencies. Moreover, this network also presents a strong multi-time-scale representative within the proposed architecture, while GRUs can exploit temporal information at less computational cost than the popular long-short-term memory (LSTM). The novelty of the proposed architecture is to employ multiple convolutional-based branches with different filter lengths to learn multi-time-scale features jointly, accelerate the learning process, and reduce overfitting by leveraging shared representations as the auxiliary information. This study also compares the multi-branch ResAttGRU networks with state-of-the-art deep learning methods using 18 years of NSRDB data at 12 solar sites. Finally, the proposed multi-branch ResAttGRU requires 7.1% fewer parameters than multi-branch residual LSTM (ResLSTM) while achieving similar average RMSE, MAE, and R-squared values.

This publication has 33 references indexed in Scilit: