A Novel Framework for Solar Irradiance Prediction Integrating Signal Decomposition With Hybrid Time-Series Models

This study aims to improve the accuracy of solar irradiance forecasting by integrating signal decomposition techniques with deep learning and statistical models. Three decomposition methods—EMD, EEMD, and CEEMDAN—were combined with models such as LSTM, Bi-LSTM, GRU, Transformer, and statistical models such as ARIMA, ARCH, and GARCH were also implemented. Hybrid model architecture was developed to process high and low frequency signals separately in this research, respectively. Among all configurations, the EEMD–Transformer–GRU model achieved the best performance, with a MAPE of 7.87%, reducing prediction error by approximately 57% compared to non-decomposed models. These findings demonstrate that the integration of signal decomposition and hybrid deep learning significantly improves forecasting accuracy and offers a promising approach for renewable energy applications.

Abstract licence: CC BY-NC-ND

40 more references, whose wording the paper’s licence keeps private.

A person counts once a day per paper. We keep no cookie and no address: a reader is a hash that changes every day. Any month under five reads shows no number.