Networked Control of PEM Fuel Cells Using an Adaptive Predictive Control Approach

In this paper, the problem of networked control of PEM fuel cells is illustrated. The extended Kalman filter and the singular pencil model methods are used to simultaneously estimate the parameters, states, and noise information for the state feedback predictive controller design. Using a networked controller developed by our research group, a closed-loop networked control system is established for the PEM fuel cell. The control performance of the closed-loop networked control fuel cell is demonstrated by practical experiments.