New modified Smith predictor scheme for integrating and unstable processes with time delay

Abstract

The paper extends a recently published Smith predictor scheme for controlling integrating and unstable processes with time delay, which results in a simpler control structure and improved tuning capacities for reference inputs and load disturbances. In the proposed scheme, a proportional controller is subtly employed to stabilise the setpoint response, and then an H2 optimal controller is analytically designed for setpoint tracking. To obtain the integral-squared-error (ISE) performance objective, a practically desired disturbance rejection transfer function is proposed to design the disturbance estimator in the inner closed loop of the proposed control structure. Essentially, the proposed control structure gives two-degrees-of-freedom (2DOF) control and correspondingly the setpoint and load disturbance responses can each be tuned conveniently by a single control parameter. Hence the setpoint response is decoupled from the load disturbance response, and both can be quantitatively estimated in terms of the proposed analytical tuning procedures. Simulation examples are included to show the superiority of the proposed method.