Fuel Efficient Balance of Plant and Power Split Control Strategies for Fuel Cell Vehicles

Abstract

Fuel cell (FC) systems with on-board hydrogen storage offer long range, fast refueling, with low audible and thermal signatures. These attributes make fuel cell vehicles (FCVs) the best option for fleet vehicles with high utilization and stringent environmental requirements. A FC system consists of the stack, which performs the electrical conversion of hydrogen to electricity, the balance of plant (BOP) components including pumps, ejectors, and blowers which are responsible for supplying reactants (hydrogen and air) at the correct rates, humidification, and thermal management hardware. Fuel cells (FCs) are typically hybridized with a battery to recuperate the braking energy and improve the system durability by reducing the a) transient and high current spikes, b) idling conditions with high open circuit potential, and c) the number of startup and shutdown cycles. At the vehicle level satisfying driver's torque/power demand is achieved by choosing the power split between the fuel cell and battery. Low hydrogen consumption and vehicle efficiency can be achieved through load preview and simultaneous optimization of vehicle speed and power split to regulate battery state of charge and fuel cell thermal management as this thesis shows.

This dissertation presents control strategies to address the above challenges for different size and weight of fuel cell vehicles motivated by the diversity of powertrains managed in the defense industry. Air-cooled stacks are considered for small power systems such as ground robots. To this end, an air-cooled fuel cell system model with a fan as a BOP component is considered. The optimization of the lumped thermal dynamic addresses the FC bulk temperature taking into account the parasitic loss of the electric fan that supplies air for the reaction and cooling simultaneously. We analyze prior work that used an offline numerical optimization method called General Purpose Optimal Control Software (GPOPS) to solve the optimal fan flow and fuel cell current for this combined BOP and powersplit optimization strategy. We show that the optimal FC temperature and current setpoints depend on the drive cycle, but their values does not change substantially within the cycle. Given the intra-cycle invariance of the setpoints, we develop two proportional-integral (PI) controllers to achieve the power split and the BOP.

Secondly, a large fuel cell vehicle (FCV) with multiple kW of power uses a liquid cooling hardware strategy and imposes low parasitic losses hence the optimization emphasis shifts on the power split strategy. A dynamic programming and equivalent minimization consumption control strategies are developed and compared for different battery sizes and battery cell chemistries.

Thirdly, co-optimization and sequential optimization of the velocity profile and the power split were compared and developed for a liquid-cooled FC with battery for a Small Multipurpose Equipment Transport (SMET) vehicle in terms of energy consumption, operating modes, and computational cost. Last but not least, the importance of accurate SOC estimation for the battery utilization is addressed by combining voltage and force measurements to improve SOC estimation for an efficient, scalable, and safe fuel cell vehicle system.