eprintid: 10163769 rev_number: 7 eprint_status: archive userid: 699 dir: disk0/10/16/37/69 datestamp: 2023-01-26 14:35:56 lastmod: 2023-01-26 14:35:56 status_changed: 2023-01-26 14:35:56 type: article metadata_visibility: show sword_depositor: 699 creators_name: Yuan, X creators_name: Chen, F creators_name: Xia, Z creators_name: Zhuang, L creators_name: Jiao, K creators_name: Peng, Z creators_name: Wang, B creators_name: Bucknall, R creators_name: Yearwood, K creators_name: Hou, Z title: A novel feature susceptibility approach for a PEMFC control system based on an improved XGBoost-Boruta algorithm ispublished: pub divisions: UCL divisions: B04 divisions: C05 divisions: F45 note: © 2023 The Authors. Published by Elsevier Ltd. under a Creative Commons license (https://creativecommons.org/licenses/by-nc-nd/4.0/). abstract: Data-driven modelling methods are being developed in the quest to achieve more accurate performance prediction of protons exchange membrane fuel cell (PEMFC) systems in response to their complicated physicochemical phenomena. However, there is little research in this field detailing the pre-processing and selection of balance of plants (BOP) features for the input layer of system performance prediction at different current densities. Furthermore, most of the previous research applies neural networks based on simulation data rather than real-time bench or vehicle operation datasets which leads to low robustness and unreliable practical results. This paper details the application of a novel algorithm denoted XGBoost-Boruta, which utilises the combination of an ensemble learning approach and a wrapping approach, to improve the robustness of feature selection and to increase the accuracy and robustness of PEMFC system performance prediction. By introduction of the Z score and shadow features to eliminate the randomness of conventional ensemble learning methods, seven key controllable BOP variables of the hydrogen anode, air cathode and cooling subsystems are selected as the original input variables to determine their dependency on the stack voltage. Two case studies are presented for verification and validation of the proposed algorithm based on the real-time dataset of bench experimental data and data obtained from heavy truck operation at current densities ranging from 100 to 1500 mA/cm2. The feature selection strategy, based on the proposed XGBoost-Boruta algorithm, largely decreases the RMSE by 23.8% and 14.1% and the R2 increases by 0.06 and 0.04 of both the bench experimental and the heavy truck validation datasets respectively. date: 2023-04 date_type: published publisher: Elsevier BV official_url: https://doi.org/10.1016/j.egyai.2023.100229 oa_status: green full_text_type: pub language: eng primo: open primo_central: open_green verified: verified_manual elements_id: 2002066 doi: 10.1016/j.egyai.2023.100229 lyricists_name: Bucknall, Richard lyricists_id: RBUCK26 actors_name: Halliday, Neil actors_name: Flynn, Bernadette actors_id: NAHAL77 actors_id: BFFLY94 actors_role: owner actors_role: impersonator full_text_status: public publication: Energy and AI volume: 12 article_number: 100229 citation: Yuan, X; Chen, F; Xia, Z; Zhuang, L; Jiao, K; Peng, Z; Wang, B; ... Hou, Z; + view all <#> Yuan, X; Chen, F; Xia, Z; Zhuang, L; Jiao, K; Peng, Z; Wang, B; Bucknall, R; Yearwood, K; Hou, Z; - view fewer <#> (2023) A novel feature susceptibility approach for a PEMFC control system based on an improved XGBoost-Boruta algorithm. Energy and AI , 12 , Article 100229. 10.1016/j.egyai.2023.100229 <https://doi.org/10.1016/j.egyai.2023.100229>. Green open access document_url: https://discovery.ucl.ac.uk/id/eprint/10163769/1/1-s2.0-S2666546823000010-main.pdf