Calculating setpoint range under actuator saturation and stochastic disturbance

Minimum variance (MV) has been widely used as a benchmark to assess the regulatory performance of control loops. However, due to the physical constraints (saturation) of the actuator, the MV performance cannot be achieved in many cases. In this paper, we investigate the output regulatory performance achievable under actuator saturation. The constrained minimum variance controller is first introduced. The performance of the linear LQG controller under saturation is also discussed. A procedure on how to compute the maximal achievable set-point range under saturation limits is presented. It is believed that with the information on achievable set-point range, the high level optimization can be greatly simplified