Transient analysis of a solid oxide fuel cell unit with reforming and water-shift reaction and the building of neural network model for rapid prediction in electrical and thermal performance

Intelligence Techniques in Sustainable Energy: Analysis of a Decade of Advances

“…Echo-state neural networks have also emerged as an effective tool for predicting degradation [143]. Specialized neural network models, such as the wavelet transform combined with long short-term memory (LSTM) and gradient boosting decision tree (GBDT), have achieved exceptional results in various facets of fuel cell prediction [146][147][148]. Techniques like merging convolutional neural networks (CNNs) with random forest feature selection and spatiotemporal vision-based deep neural networks with 3D inception LSTM have shown significant advances in fuel cell vehicle speed predictions [145,149].…”