Automatica, Vol.30, No.2, 277-292, 1994
Near-Optimal Control of Nonstandard Singularly Perturbed Systems
Near-optimal control of nonstandard multiparameter/multitime scale singularly perturbed systems is studied using the descriptor variable approach. In particular, the slow models of the considered systems are allowed to exhibit impulses. In the case of multiparameter singular perturbations, the near optimal control is a cascade connection of separately designed slow and fast subregulators; in the case of multitime scale singular perturbations, the near-optimal control is hierarchically composed of N + 1 subregulators, a parallel algorithm is provided for designing the subregulators separately. The knowledge of the small singular perturbation parameters is not required in the overall design. Furthermore, the frequency responses of the near optimal systems are analyzed, and several design examples are given.