Experimental study of water management of a three-cell PEM fuel cell stack using adaptive neuro-fuzzy approximation and fuzzy control

• Published in: International journal of hydrogen energy 2024-01, Vol.50, p.931-944
• Main Authors: Khanafari, Ali, Alasty, Aria, Kermani, Mohammad Jafar, Asghari, Saeed
• Format: Article
• Language: English
• Subjects: Adaptive neuro-fuzzy; Dehydration; Flooding; Fuzzy; Polymer electrolyte membrane fuel cell (PEMFC)
• ISSN: 0360-3199; 1879-3487
• DOI: 10.1016/j.ijhydene.2023.09.192

Day by day, polymer electrolyte membrane fuel cell (PEMFC) finds a growing place as a clean energy converter. Identifying and controlling operational defects can increase its useful life and reliability. The present study addresses a countermeasure against stack flooding and dehydration as a continuation of the current authors' previous work. Three mechanisms (i.e., controlling the stack temperature and flow of both sides) applying fuzzy logic manage stack moisture content. Three fuzzy controllers use voltage and cathod&anode pressure-drops dimensionless deviations from ANFIS models using the previous study calculated thresholds (0.0092 bar, 0.0047 bar, and 0.0735 V). Using the method reduces the cathode pressure drop positive deviation (a characteristic of flooding) by 61% and reduces the voltage deviation by 48% and 47% in deliberated flooding and dehydration experiments, respectively. In other words, using the method returns the status closer to healthy conditions, which indicates successful PEMFC water management. •Flooding & dehydration experiments of a three-cell PEM fuel cell are performed.•Three ANFIS models are used to diagnose cathode/anode flooding and dehydration.•Three mechanisms are proposed to decrease or overcome flooding & dehydration.•The fuzzy controllers efficiently manage three proposed mechanisms.•The method reduces the flooding & dehydration effect on voltage reduction up to 48%.