For high frequency vibration rejection, constant boundary layer control design and dynamic sliding mode control (DSMC) design using an integrator have been proposed. However, they use only one output feedback, so disturbance rejection and robustness enhancement are limited.
Adaptive SMC (ASMC) is used for control of systems with unknown or bounded disturbances. However, these methods do not keep the sliding variable at the ideal value zero. To increase the robustness of Smith predictor for control of delay processes, a robust Smith predictor and a PI control method for multivariable processes have been developed.
Kim Ho, dean of the Faculty of Automatics, developed a linear PI adaptive dynamic sliding mode control (DSMC) system based on Robust Smith Predictor (RSP).
In order to primarily suppress the disturbance and model perturbation of a plant, he implemented the equivalent control by a PI output feedback control loop. As the outer switching controller is designed for extended systems with an integrator, it automatically becomes a DSMC and chattering free. A linear PI feedback system with sliding variables as output is used for switching gain adaptation of sliding-mode control, and consequently, the optimality and adaptability of PI adaptive algorithm are guaranteed by the linear feedback control theory.
He verified through simulation that the proposed SMC system shows high performance that cannot be achieved by any one control alone for all indexes, namely, tracking performance, disturbance rejection, robustness, adaptability and chattering rejection.
For more information, please refer to his paper “Decentralized PI Adaptive Sliding Mode Controller with Double Output Feedback for TITO Processes” in “Proceedings of KUTIC-2025”.