Performance Evaluation of Direct Model Reference Adaptive Control on a Coupled-tank Liquid Level System

This paper presents the performance evaluation on the application of direct model reference adaptive control (DMRAC) with various types of command inputs in a process plant. A DMRAC algorithm with second order reference model scheme is proposed for the system. The coupled-tank liquid level control system is used as a process plant test bed. A simulation is carried out using MATLAB and Simulink to control the modelled nonlinear coupled tank under various experimental conditions. Different forms of commanded inputs such as of square wave, sinusoidal and triangle (ramp) input signal are fed to the system controlled by DMRAC with different reference models that specify performance criteria. The controller is robust to produce the appropriate control signals in response to the desired transient performance defined by reference model to control the liquid level of the coupled-tank in the presence of plant nonlinearity, plant parameter changes and measurement noise