Performance Prediction of Alkaline Fuel Cell Via Gradient Boosting Algorithm

Abstract

With the DFT theory at the forefront, non-PGM catalysts have begun to develop, and in much literature, results have shown that the performance of non-PGM catalysts is comparable to that of PGM catalysts. Jing et. al. synthesized Fe-Co dual sites on nitrogen-doped carbon for oxygen reduction reaction (ORR). (FeCo)/N-C had almost equal half-wave potential to commercial Pt/C.[1] Hanguang et. al. achieved high ORR-activated Fe-N-C catalysts by tuning the doped Fe content and active sites.[2] However, if both (FeCo)/N-C and Fe-N-C were applied to PEMFC, they were not comparable to PGM catalysts unlike in the results of the half-cell test. This discrepancy means that numerous factors should be considered and optimized in a practical system like a fuel cell. Besides, it is indicated that even if the DFT theory and deep-learning combine to suggest catalysts with high performance, it may not produce as much performance in the actual fuel cell system as expected.In this respect, the combination of fuel cell systems (not catalysts) and artificial intelligence (AI) seems more reasonable. Without human judgment or intervention, machine learning sets priorities between each index by learning from many databases.Herein, we newly developed an optimization model for alkaline liquid fuel cell (Hydrazine fuel cell) using gradient boosting algorithm (XGBoost) which is one of machine learning algorithms. We operated fuel cells in various conditions by changing humidity of cathode, back pressure of cathode, cell temperature, stoichiometric factor (air/fuel), and concentration of the fuel. And then, we categorized and classified with the specific algorithm. Finally, we re-organized and set as a function of weight which effects on fuel cell operation. We hope that these approach will help improve fuel cell performance by controlling a number of factors without human intervention.[1] J. Wang, Z. Huang, W. Liu, C. Chang, H. Tang, Z. Li, W. Chen, C. Jia, T. Yao, S. Wei, Y. Wu, Y. Li, J. Am. Chem. Soc. 139 (2017) 17281–17284.[2] H. Zhang, H.T. Chung, D.A. Cullen, S. Wagner, U.I. Kramm, K.L. More, P. Zelenay, G. Wu, Energy Environ. Sci. 12 (2019) 2548–2558.