Analytical solution and experimental validation of the energy management problem for fuel cell hybrid vehicles

The objective of an energy management strategy for fuel cell hybrid propulsion systems is to minimize the fuel needed to provide the required power demand. This minimization is defined as an optimization problem. Methods such as dynamic programming numerically solve this optimization problem. Strategies such as the equivalent consumption minimization strategy derive an analytical solution based on low-order models that approximate fuel cell stack and battery behavior. This paper presents an analytical solution based on models of the fuel cell system and battery close to physics. Apart from an analytical solution, this solution provides a fundamental understanding of the energy management problem. Because the solution is analytic and does not need a priori knowledge, the computation time is limited, and real-time implementation is possible. The solution presented is validated against existing optimizing energy management strategies in both simulations and experiments. For simulations, a midsize distribution truck is chosen. Experiments are carried out on a 10-kW scale test facility that comprises a fuel cell system, a battery, a motor with load, and an electronic load. In both simulations and measurements, the solution presented in this paper performs best compared to the equivalent consumption minimization strategy and a range-extender strategy, although the differences are within 3%. In the simulations, the solution presented approaches a minimum in fuel consumption, derived offline using dynamic programming, within 1%