A self-optimizing switching control scheme for uncertain ARMAX systems

In this paper, we propose a switching control scheme for uncertain AutoRegressive Moving Average eXogenous (ARMAX) systems, which guarantees stability of the switched control system and self-optimality. The adopted switching control logic is based on the Extended Least Squares (ELS) parameter estimation procedure for controller selection and switching. An asymptotically vanishing dither noise is added to the control input so as to introduce appropriate excitation of the system dynamics for consistently estimating the system parameters.