Power management optimization of fuel cell/battery hybrid vehicles with experimental validation

Fuel cell hybrid vehicles offer a high-efficiency and low-emission substitute for their internal combustion engine counterparts. The hybridization significantly improves the fuel economy of the vehicle; however, exploiting the hybridization requires a well-designed power management strategy that optimally shares the power demand between the power sources. This paper deals with the optimization of power management strategy of a fuel cell/battery hybrid vehicle, both off-line and in real-time. A new formulation of the optimization problem for the real-time strategy is presented. The new approach allows the optimization of the controller over a set of driving cycles at once, which improves the robustness of the designed strategy. The real-time optimization is applied to two forms of real-time controllers: a PI controller based on Pontryagin's Minimum Principle with three parameters and a fuzzy controller with ten parameters. The results show that the PI controller can outperform the fuzzy controller, even though it has fewer parameters. The real-time controllers are designed by simulation and then validated by experiment.