Sherpa is strongly implied in the network design and in the choice of the actuators (air compressor, radiator, valve …).
Our contribution consists in validating the components choices using a dynamic environmental model of the system (using j‑Graph methodology). This model, based on the thermo-fluid equations of the system, takes into account the characteristics of each part and helps to validate the requirements of the system.
The control system includes the basic control laws, diagnosis and monitoring as well as the global supervisor of the system.
Sherpa Engineering develops a model based control system of the entire fuel cell system including the fuel cell stack and three main subsystems :
- air subsystem : supply of the process air at the required pressure, flow rate, temperature, and humidity
- hydrogen subsystem : supply of hydrogen at the required pressure and flow rate
- cooling subsystem : guarantee adequate cooling of fuel cell stacks and ensure small temperature gradient across stacks
The model based control design approach consists on using a hierarchical decomposition of the system and, on the other hand, on using a model based predictive control methodology. MBPC allows handle in a straight-forward way the multivariable interactive control problems.
This modular and efficient approach is easily adaptable to different sorts of fuel cell systems or even to similar traditional systems (thermal management of a classic thermal engine ...).
Diagnosis of a fuel cell system
The focus of the diagnosis system is to avoid any deterioration of the fuel cell stack and to increase its lifespan.
Sherpa Engineering proposes a simple but also efficient diagnosis-oriented model of the fuel cell system. It includes three main monitoring modules :
- Cell voltage monitoring
- Vital parameters for safe and proper operation : membrane temperature, air flow rate, stoichiometric ratio, hydrogen and air pressure stack inlet, pressure differential …
- Actuators and sensors diagnosis
The fault diagnosis system is based on a reduced fuel cell model. A fault detection and identification module is responsible for processing a residual (difference between the predicted and actual sensor reading) to decide which fault has occurred.
Detection of hydrogen leaks
Another common safety concern for fuel cell systems are hydrogen leaks which are potentially dangerous faults.
In addition to hydrogen sensors, we develops a method for detecting leaks in a fuel cell system that includes a hydrogen storage tank, a primary shut-off valve and a secondary shut-off valve in a supply line.
Validation
Validation process :
- Fast Prototyping with Simulink and dSPACE systems
- Hierarchical decomposition : using partial bench to validate subsystems before the final validation on the entire fuel cell system
Model based design : by running the same tests on the model, we know exactly how the design should perform in the lab or with the vehicleIndustrial
Applications
FISYPAC is a partnership project which intends to develop a fuel cell system and to test it on a fuel cell electric vehicle.
Sherpa Engineering is in charge of the design and the validation of the control and diagnosis system and of the energy management strategy.
The hydrogen fuel cell puts out 20 kW, giving the car a 350 km range and a top speed of 130 km/h.