Model-Based Development of the “Big 5” in Fuel Cell Systems

Fuel cell systems are gaining traction across heavy-duty applications, driven by global decarbonization targets. Managing their inherent complexity and diverse architectural requirements, commonly organized into the “Big 5” fuel cell subsystems (stack, thermal, electric, anode, and cathode), necessitates advanced Model-Based Development (MBD) approaches. This paper presents and validates a constraint-graph-based, equation-oriented, acausal MBD methodology for fuel cell system (FCS) development, implemented in an industrial modeling environment. This methodology supports scalable functional and software development from 75 kW single-stack systems to twin-stack configurations exceeding 250 kW. It facilitates robust parameterization and reuse of consistently formulated, subsystem-level physical models across Model-in-the-Loop (MiL) to Hardware-in-the-Loop (HiL) environments, ensuring numerically robust software architectures and improved embedded control quality. Industrial application has demonstrated workflow benefits, including indicative reductions of approximately 30 percent in development time, 30 percent in calibration effort, and up to 15 percent in ECU memory utilization across multiple fuel cell system development programs. These improvements are primarily achieved through modular model formulation, systematic model reuse, and elimination of artificial delays enabled by the acausal modeling framework.