Controller compact form dynamic linearization based model free adaptive control

Abstract

A new type of model free adaptive control (MFAC) method, by virtue of the compact-form-dynamic-linearization based controller (CFDLc), is presented for a class of discrete time SISO nonlinear systems. The proposed method is a pure data-driven control (DDC) method since the controller is designed using measured I/O data of the controlled plant rather than first principles model of the controlled plant. This method contrast with model-based control method mainly in two fundamental aspects: it is not based on the knowledge of a process model and it intend to make direct use of the information carried by the measured I/O data in order to adjust the numerical parameters of a controller whose structure is derived from the equivalent dynamic linearization data model of an ideal nonlinear controller. Different from the prototype of MFAC, the proposed method uses the dynamic linearization approach both on controller and controlled plant. The stability of the CFDLc-MFAC is guaranteed by rigorous theoretical analysis and the effectiveness is verified by simulation results.