Dynamic anti-windup design in anticipation of actuator saturation

In a traditional anti-windup design, the anti-windup mechanism is set to be activated as soon as the control signal saturates the actuator. A recent innovation of delaying the activation of the anti-windup mechanism, both static and dynamic, until the saturation reaches to a certain level of severity has led to a performance improvement of the resulting closed-loop system. More recently, it has been shown that significant further performance improvements can be obtained by activating a static anti-windup mechanism in anticipation of actuator saturation, in comparison with the delayed activation design. This paper demonstrates that anticipatory activation of a dynamic anti-windup mechanism would also lead to significant performance improvements over both the immediate and delayed activation schemes.