Hybrid adaptive control of nonlinear system with two types of external linear models

Abstract

The paper deals with the hybrid adaptive control of the water level inside the real model of the water tank. The controlled system is nonlinear because the output volumetric flow rate from the tank is nonlinear function of the water level. The adaptivity is satisfied by the recursive identification of the external model as a linear representation of the originally nonlinear system. The External Linear Model (ELM) is described by the linear transfer function in continuous-time. Delta models were used here for the identification. Although delta models belongs to the class of discrete-time models, it was proofed that their parameters approaches to the continuous-time ones and that is why we call this strategy "hybrid" - the controller is defined in the continuoustime but the identification runs in discrete-time. The main task of this contribution is to compare the simulation results for the first and the second order transfer functions as a ELM of this system.