Design and applications of an optimal anti-windup digital controller using scalar sign function approach

This paper presents a scalar sign function based anti-windup control method for the sampled-data system with input saturation. It is found that input saturation can be effectively approximated by the smooth scalar sign function. When saturation encounters, an iteration procedure is executed in which the weighting matrices are systematically adjusted in the LQR performance index until the obtained control inputs do not violate the limits. The adjustments are based upon computations centered on the utilized scalar sign function. A general digital control scheme is described in which optimal linearization and Chebyshev quadrature digital redesign method are applied. A simulation example and a real time experiment are shown to demonstrate the effectiveness of the proposed digital controller.