| Hydrogen polymer electrolyte membrane (PEM) fuel cells are the fuel cell of choice for the automotive industry. The problems addressed in this work are the design and modelling of a hybrid fuel cell power system and the optimization of its fuel efficiency, online, while protecting the fuel cell health during transients.; To help meet the transient demands on fuel cell propelled automobiles, ultracapacitors are proposed as auxiliary power devices, due to their fast discharge dynamics. In simulation, ultracapacitors mitigate the disruptive behavior in the system caused by the faster parasitic and load electrical dynamics as compared to the fuel cell power delivery system. This helps to prevent potentially permanent damage to the fuel cell membrane. To improve performance further, an optimization problem is proposed for the hybrid power system. The goal of the optimization is to minimize hydrogen fuel usage, online, at a given load disturbance, established by the driver.; Since accurate modelling of fuel cell dynamics is an open problem, an extremum-seeking control method is used in this work. This method does not rely on a preexisting model of the system dynamics. It adaptively updates a basis function, online, representing the optimization curve between the input and output. The controller simultaneously updates the control input to find the point of zero gradient, or the optimum. In simulations, the controller can find the optimum operating point for the fuel cell power system, at a given load level. |
