Model-free adaptive fixed-time prescribed performance control for PEMFC air feeding system based on oxygen excess ratio reconstruction.

The polymer electrolyte membrane fuel cells (PEMFCs) systems with unmeasurable variables, output constraints, time-varying disturbances, and uncertain dynamics can hardly be modelled well, such that model-based control strategies become infeasible. Therefore, an ultra-local model (ULM) only utilizing input and output signals of the PEMFC system is innovatively presented. Together with a fixed-time disturbance observer (FxTDO), fixed-time prescribed performance control, and an adaptive technique-based compensator, a novel internal state estimator-based model-free adaptive fixed-time prescribed performance control (MF-AFxTPPC) is naturally developed to estimate and regulate the oxygen excess ratio (OER). Main advantages are presented as follows: (1) The proposed internal state estimator is designed to estimate OER via FxTDO, which is known for its fixed-time convergence and low computational time. (2) The presented ULM algorithm not only reduces the complexity of controller design but also significantly enhances its flexibility and adaptability. (3) The FxTPPC scheme constrains the tracking error and stabilizes the closed-loop system, which is more favorable to practical implementations. (4) The adaptive technique-based compensator can not only compensate for estimation error and mitigate the input chattering but also guarantee the global robustness. Furthermore, the Lyapunov theorem is utilized to analyze the stability of the designed MF-AFxTPPC scheme. Finally, the numerical simulations on a nonlinear PEMFC model with the proposed controller strategy are achieved in a MATLAB/SIMULINK environment, and findings are given to exhibit the robustness and superiority of the presented method.