Enhanced Smith predictor for unstable processes with time delay

In this paper, a modified form of Smith predictor is proposed for unstable processes with time delay. A Smith predictor structure is incorporated separately for servo and regulatory problems, and an additional controller is added to improve disturbance rejection. Analytical formulas are given for the design of controllers. The proposed method has only one tuning parameter. A first-order filter is used to achieve good robustness for uncertainties in process parameters. The designed controllers give good set-point tracking and disturbance rejection. Quantitative improvements are given and compared to the previous methods. Illustrative examples are given to show the advantages of the proposed method.