Premium
System behavior prediction by artificial neural network algorithm of a methanol steam reformer for polymer electrolyte fuel cell stack use
Author(s) -
Qi Yuanxin,
Andersson Martin,
Wang Lei,
Wang Jingyu
Publication year - 2021
Publication title -
fuel cells
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.485
H-Index - 69
eISSN - 1615-6854
pISSN - 1615-6846
DOI - 10.1002/fuce.202100006
Subject(s) - stack (abstract data type) , steam reforming , electrolyte , artificial neural network , computer science , algorithm , proton exchange membrane fuel cell , methanol , backpropagation , relation (database) , hydrogen production , hydrogen , fuel cells , process engineering , chemical engineering , materials science , engineering , chemistry , electrode , artificial intelligence , organic chemistry , database , programming language
In this paper, a novel membrane reactor (MR) for methanol steam reforming is modeled to produce fuel cell grade hydrogen, which can be used as the inlet fuel for a later developed 500‐W horizon polymer electrolyte fuel cell (PEFC) stack. The backpropagation (BP) neural network algorithm is employed to develop the mapping relation model between the MR's prime operational parameters and fuel cell output performance for future integration system design and control application. Simulation results showed that the MR model performs well for hydrogen production and the developed PEFC system presents good agreement with experimental results. Finally, the BP method captures an accurate mapping relation model between the MR inputs and PEFC output, for example, predicts the system's behavior well.