Degradation-Conscious Control for PEM Fuel Cell Systems

Abstract

Cost and durability remain the main roadblocks to widespread commercialization of polymer electrolyte membrane (PEM) fuel cell technology for automotive applications. Significant research efforts over the past decade have resulted in considerable improvements on both fronts. Most of this research has focused on developing new materials for enhanced durability. Nevertheless, as demonstrated by the commercialized Toyota Mirai, system-level strategies designed to effectively utilize the full potential of a given fuel cell stack also hold a great promise in realizing the durability targets. Additionally, such system-level approaches can reduce the overall cost, for example by downsizing or eliminating the humidifier. Despite their potential, these system-level mitigation strategies are not commonly pursued in the literature. As a starting step towards developing a system-level strategy to mitigate membrane degradation, in this talk we present a degradation-conscious control algorithm that aims to meet the requested power demand while respecting the degradation constraints to ensure safe operation of the fuel cell. This is achieved using the model-predictive-control (MPC) framework [1]. In particular, we present a coarsely discretized version of a two-phase and non-isothermal model for transport phenomena in the through-the-membrane direction [2], [3]. We then show how the model can be simplified for the purpose of the control design. To ensure feasibility of real-time control law generation, we further simplify the model by linearizing it about the current operating point. Having derived the linear model, we show how we can construct a degradation-conscious controller that respects the bounds on the membrane hydration and its temperature using the MPC framework. Our preliminary results indicate that the approach can be used to control the state of the membrane in order to ensure its safe operation [4]. [1] J. B. Rawlings and D. Q. Mayne, Model predictive control: Theory and design. Nob Hill Pub. Madison, Wisconsin, 2009. [2] A. Goshtasbi, B. L. Pence, and T. Ersal, “Computationally efficient pseudo-2D non-isothermal modeling of polymer electrolyte membrane fuel cells with two-phase phenomena,” J. Electrochem. Soc., vol. 163, no. 13, pp. F1412–F1432, 2016. [3] A. Goshtasbi, B. L. Pence, and T. Ersal, “A real-time pseudo-2D bi-domain model of PEM fuel cells for automotive applications,” in ASME 2017 Dynamic Systems and Control Conference, 2017, p. V001T25A001-V001T25A001. [4] A. Goshtasbi and T. Ersal, “LQ-MPC Design for Degradation-Conscious Control of PEM Fuel Cells,” in Proceedings of the 2019 American Control Conference, 2019.