Deep learning solar forecasting for green hydrogen production in India: A case study

Solar energy based green hydrogen production is dependent on energy produced from Photovoltaic (PV) panels that is in turn dependent on Global Horizontal Irradiance (GHI), which has stochastic and intermittent nature. The intermittent power output from PV panels may interfere with the steady production of electricity required for electrolysis process to synthesis the green hydrogen thereby; reducing the effectiveness and scalability of green hydrogen generation. Ultimate goal of this work will be to effectively forecast and develop the atlas map for solar based green hydrogen production potential using the proposed algorithm i.e. Complete Ensemble Empirical Mode Decomposition with Adaptive Noise-Bidirectional Long Short-Term Memory (CEEMDAN-BiDLSTM). In order to suggest a reliable and precise forecasting methodology that will support India's low-carbon economy goals, the proposed method is evaluated on GHI datasets acquired via national portal of National Institute of Wind Energy (NIWE) for two Indian sites i.e. Bhadla and Fatehgarh.(c) 2023 Hydrogen Energy Publications LLC. Published by Elsevier Ltd. All rights reserved.