Extended kalman filter for quantifying hydrogen leaks in PEM fuel cells by estimating oxygen concentration

Hydrogen transfer leaks are one of the most important life-limiting faults in polymer electrolyte membrane fuel cells (PEMFCs). Hydrogen transfer leaks result in a reduction in the amount of oxygen available in the cathode (air) channel, with large leaks resulting in oxygen starvation and hydrogen emission. This paper aims to develop an adaptive extended Kalman filter (EKF) to estimate the unknown oxygen concentration, which is then used to infer hydrogen leaks. To this end, the paper first develops the lumped model of the fuel cell from a pseudo-2D model of a fuel cell. Next, a (non-adaptive) EKF is developed to estimate the fuel cell states under both normal and oxygen-starved conditions. The adaptive EKF is then implemented by adding the unknown hydrogen leak to the list of estimated states. Finally, the paper demonstrates the efficacy of the proposed adaptive EKF by using it to accurately estimate unknown hydrogen leaks in a high-fidelity virtual fuel cell under excessively noisy conditions.