On-line Degradation Estimation of Proton Exchange Membrane Fuel Cell based on IMM-EKF

As a new alternative energy device, Proton Exchange Membrane Fuel Cell (PEMFC) has wide application potential in many fields such as electric vehicles and digital generators. In this paper, an on-line recursive filtering approach is presented to estimate the performance degradation of the PEMFC based on Interacting Multiple Model (IMM) and Extended Kalman Filtering (EKF). The change degree of the equivalent resistance and maximum current density of the PEMFC is used as the indicator of performance degradation. A semi-empirical model is introduced to establish the relationship between the monitoring voltage and the degradation indicator. Then, two nonlinear filtering models are constructed and the IMM-EKF algorithm is used to update the estimate of the degradation indicator with real-time measurements. Finally, the proposed approach is validated by an experimental dataset.