Anti-windup for time-varying delayed cellular neural networks subject to input saturation

This paper deals with the problem of anti-windup design for a class of state saturation systems subject to time-varying delayed cellular neural networks and input saturation. By introducing the saturation degree function and applying the convex hull theory to handle the saturated terms, we firstly put forward a stabilization controller for the time-varying delayed system in the absence of input saturation via LMI formulation according to Lyapunov-Krasovskii theorem. Then the anti-windup gain matrix is derived to compensate for the difference between the constrained and unconstrained systems in the presence of input saturation. Further, the enlargement to the basin of attraction under input saturation is formulated, and the corresponding optimization problem with LMI constraints is given. Finally, numerical examples are included to illustrate the effectiveness of the proposed design technique.