Decoupling Control of the Quadruple-Tank System

Level control of a quadruple-tank process with variable nonlinear zero dynamics is investigated in this paper. The plant provides a nice benchmark problem that illustrates some of the challenges posed by complex interconnected systems for linear design paradigms. The system has some special features like nonlinear dynamics, multi-input multi-output, and an adjustable zero location that makes it ideal for illustrating many concepts in multivariable control. Most of the recently proposed control methods suffer from a nice decoupling control scheme. Among these control schemes model predictive control can decouple the system if an accurate model of the system is used in the control design stage. In this paper an adaptive nonlinear feedback linearization scheme is proposed that can completely decouple and linearize the system in the all working points even in the presence of a model-mismatch. A comparison between the proposed controller, conventional PID control and the constrained model predictive control scheme is also given by some simulation results