Current Distribution Identification in Fuel Cell Stacks From External Magnetic Field Measurements

This paper presents a new non invasive technique for fuel cell diagnosis. The method relies on the measurements of the magnetic field signature generated by a working fuel cell. Knowing the relationship between currents and magnetic field, it is possible to estimate the current density by solving an inverse problem. This problem being ill-posed, original current and sensors basis are proposed to improve the reconstruction process. The experimental set-up remains simple, based on few fixed sensors and shows a good efficiency.