A modified Smith predictor based – Sliding mode control approach for integrating processes with dead time

Abstract

This paper proposes a modified hybrid robust Smith predictor controller design for integrating processes with long dead time. The proposal results in a dynamic sliding mode controller for integrating systems on blending sliding mode control and Smith Predictor modification concepts. The design part of the controller corresponding to the sliding mode arises from an approximate model of the high-order nonlinear integrating process to characterize the non-self-regulating processes. The approximation to a First Order Integrating System Plus Dead Time is obtained from an experimental method that adopts identification procedures and avoids complex analytical resources. The model contributes to designing the dynamic sliding mode controller and tuning parameters. The resulting controller is a hybrid scheme that combines the sliding mode methodology, a smith predictor structure, and a PD compensator. The hybrid control topology is used for the first time for integrating systems with dead time. The proposed approach is applied to a high order linear integrating system with long-dead time and a coupled tanks system using hardware in the loop (HIL) simulation. The hybrid control topology provides enhanced controller performance, guarantees proven robustness to parameter modeling uncertainties, and attenuates high-frequency oscillations, counteracting the adverse effects of chattering. Integral of Squared Error (ISE), Total Variation of control effort (TVu), Maximum overshoot, and Settling Time are used for quantitative comparative performance analysis.