Robust nonlinear adaptive pressure control of polymer electrolyte membrane fuel cells considering sensor failures based on perturbation compensation

The large deviations between the partial pressure of oxygen and hydrogen can lead to severe membrane damage, which reduces the life of the polymer electrolyte membrane fuel cell (PEMFC). The deviation of pressure can be restrained by effective control of partial pressure. Considering the actual operating environment of PEMFC, its control system is not only affected by load changes, but also affected by some unknown disturbances, such as system parameter variations, failures, etc. Considering the above factors, this paper has proposed a robust nonlinear adaptive pressure control (RNAPC) employing the perturbation estimation technique to suppress the deviations. The proposed controller does not need to identify or estimate the system parameters in real time like the conventional controller based on parameter estimation technology, nor does it need to rely on the fault diagnosis and isolation technology to detect and estimate the failures in real time. It only uses the variations of system state variables caused by failures to provide the required failure information for the controller. In the design of RNAPC, an observer is designed to estimate a perturbation term containing system uncertainties, nonlinearities, and disturbances. The estimation of perturbation is used for the compensation of actual perturbation. Then, an adaptive FLC of PEMFC system can be realized. Therefore, the RNAPC requires only a few state measurements and is independent of an accurate system model. Simulation results show that the RNAPC method has smaller deviations of partial pressure of oxygen and hydrogen and better dynamic performance than traditional PI control under load disturbance, and obtains higher robustness against uncertainties and sensor failures than traditional PI control and FLC. The maximum error of partial pressure can be limited within 0.5% under four different simulation scenarios under the RNAPC. (C) 2022 The Author(s). Published by Elsevier Ltd.