Feasibility study of variable gain Super-Twisting control in fuel cells based systems

This work presents a controller based on a Super-Twisting (ST) algorithm with variable gains, designed to maximize the efficiency of a fuel cell (FC). The strategy consists on regulating the oxygen excess ratio, maintaining it at an optimum value in spite of external disturbances and model uncertainties. The tested algorithm has the well-known advantages of the original ST with fixed gains, such as simplicity, robustness against uncertainty and disturbances, reduced chattering and finite time convergence. In addition, the variability of its gains allows expanding the range of operation, without a significant increase in the complexity of the control law. The adjustment of the gains is based on certain variable bounding functions, which have to be determined analysing the system and its characteristics. This process is laborious for a non-linear system like the FC, but is performed off-line during the design stage. Promising results have been obtained through exhausting simulation tests, considering model uncertainties and highly variable load regimes.