Fuel cell (FC) systems coupled with additional storage sources (batteries or supercapacitors) are a potential solution for the traction of electrified vehicles that has been considered for several years. Different studies have already addressed the issue of energy flow management between different power sources in order to optimize the global efficiency of the system. The usual approaches use a static FC model, while the physical limits of the system are integrated as optimization constraints, without taking into account the FC dynamics. The proposed control strategy explicitly considers the dynamic model of the FC to minimize the hydrogen consumption and thus achieve a maximum efficiency of the FC system. To account for the nonlinear behavior of this process, the adopted predictive control is based on a linear time-varying model (MPC-LTV). This approach uses a set of models linearized around different points belonging to a desired trajectory, consistent with the load demand of a given driving cycle. The design of the control law is performed using a quadratic criteria composed of the error between the stoichiometric coefficient of the FC and its optimum value, while guaranteeing an acceptable state of charge of the secondary storage element. For the control design, a reduced model of the FC system is used, which includes, besides the storage element, the required auxiliaries for the system operation.

Location:
Room: 12.163
Bldg: EV
1515 St. Catherine West
Montreal, Quebec
H3G2W1