A general observer based data-driven control for nonlinear discrete system

摘要

In this paper, a general observer based data-driven control algorithm is proposed to adaptive control for discrete system using input and output (I/O) data. In contrast to model based controller, data-driven control fully utilizes the I/O data measured from the object, whose accurate mathematics model is difficulty to confirm. we deploy pseudo gradient (PG) parameters of full form dynamic linearization (FFDL), which are obtained via adaptive observer based on the method model free adaptive control (MFAC). The observer proposed has more degree of freedom (DOF) and is more flexible. The tracking error is guaranteed converging to 0 and the stability analysis of the controller proposed is presented. Simulation results in MATLAB environment show the effectiveness and applicability of the control law.