Analysis of actuator saturation in feedback control systems

This dissertation examines the effects of actuator saturation on a feedback control system. Specifically, the saturation examined is not directly measured by the feedback controller and is due to a physical stop in an actuator. This lack of "visibility" of the saturation can lead to uncontrolled oscillations in the closed loop system, although the system is controllable and observable.;The failure of conventional analysis and design techniques to find or correct the saturation induced oscillations are discussed in detail. A practical method using physical means to reduce the susceptibility of the closed loop system to saturation induced oscillations is developed in detail. This practical method involves modifying the physical stop of an actuator.;Closed-loop controllability ideas are presented as a starting point for future research. A means for physical verification of the analytical tools developed in this research are described for an existing physical system, the University of Wyoming Precision Pointing and Vibration Isolation Hexapod. The University of Wyoming Hexapod is a precision Stewart platform with six mutually orthogonal actuators.;A single-input single-output system is examined in detail to assist in describing the effect of saturation induced oscillations. For this one-dimensional model a saturation induced oscillation region (SIOR) is found and described. Additionally, the effects of system parameter variation on the size and shape of the SIOR are explored.