For the first-order integrating processes with long delay, a Modified Smith Predictor (MSP) was analyzed when the control signal saturated. If the MSP controller has no anti-windup scheme, the saturation effectively c...For the first-order integrating processes with long delay, a Modified Smith Predictor (MSP) was analyzed when the control signal saturated. If the MSP controller has no anti-windup scheme, the saturation effectively chops off the control input resulting in performance deterioration. Thus, a simple and effective anti-windup scheme, without an additional parameter, is designed for the MSP. Simulation results illustrate that the proposed method may achieve good performance under the nominal and model uncertainty cases.展开更多
Uncertainty model validation has been studied for stable system by several authors. This paper considers uncertainty model validation for unstable systems. Under a similar assumption to study for stable systam, a nece...Uncertainty model validation has been studied for stable system by several authors. This paper considers uncertainty model validation for unstable systems. Under a similar assumption to study for stable systam, a necessary and suffident condition is obtained for unstable uncertainty models not to be invalidated, which generalizes the results of previous study and show that the uncertainty model validation for unstable systems is again a boundary interpolation problem.展开更多
文摘For the first-order integrating processes with long delay, a Modified Smith Predictor (MSP) was analyzed when the control signal saturated. If the MSP controller has no anti-windup scheme, the saturation effectively chops off the control input resulting in performance deterioration. Thus, a simple and effective anti-windup scheme, without an additional parameter, is designed for the MSP. Simulation results illustrate that the proposed method may achieve good performance under the nominal and model uncertainty cases.
文摘Uncertainty model validation has been studied for stable system by several authors. This paper considers uncertainty model validation for unstable systems. Under a similar assumption to study for stable systam, a necessary and suffident condition is obtained for unstable uncertainty models not to be invalidated, which generalizes the results of previous study and show that the uncertainty model validation for unstable systems is again a boundary interpolation problem.