A SLIDING MODE CONTROL APPROACH TO PREDICTIVE REGULATION

In this article, a sliding mode feedback control scheme of dynamic nature is proposed as an efficient alternative to deal robustly with the intrinsic tracking problem associated with every Model Based Predictive Control strategy. The results, which are fundamentally based on system inversion, apply to nonlinear single-input single-output perturbed systems for which the tracking of a prespecified desirable output reference signal is required. The scheme is shown to handle efficiently large modeling errors and unmatched perturbation input.