Year: 2026
Type: article
Abstract: Proton-exchange membrane fuel cells (PEMFCs) play a pivotal role in decarbonizing the transport sector due to their high efficiency and zero local emissions. This work proposes an embedded, nonlinear, two-stage model predictive control (MPC) that optimally allocates the stack supply from the air, hydrogen, and coolant subsystems. A physics-based process model of the stack captures the dominant dynamics of membrane hydration, temperature, and reactant pressures, representing key internal states that influence performance and durability. The two-stage structure separates the optimal economic planning from the real-time tracking MPC, (1) ensuring constraint-compliant regulation of the considered states under highly dynamic operation, (2) maintaining efficient stack operation, and (3) resolving the over-actuated nature of the multi-input system. Experimental validation on a full-scale PEMFC test bench demonstrates a reduction in hydrogen consumption by an average of 5.24 % per kilowatt of electric power compared to a nominal stack control. At the same time, membrane temperature and water content remain within prescribed limits.
Source: Computers & Chemical Engineering
Institution: RWTH Aachen University
Language: English
Subfield: Control and Systems Engineering
Field: Engineering
Domain: Physical Sciences
Open Access status: hybrid
Estimated APC paid: $4,050
OpenAlex ID: W7167840667