Global stabilization and disturbance suppression of a class of nonlinear systems with uncertain internal model (Q1868054)

The author considers a SISO nonlinear system which can be transformed into the output feedback form \[ \dot\zeta=A_c\zeta+bu+\varphi(y)+Dw,\qquad y=C\zeta, \tag{1} \] where \(\zeta\in\mathbb R^n\) is the state vector, \(A_c=J_n\) is the Jordan box, \(b=(0,\dots,0,b_\rho,\dots,b_n)^T\), \(u\in\mathbb R\) is the control, \(D\in\mathbb R^{n\times m}\) is a constant matrix, \(\varphi\) is a known nonlinear smooth vector field in \(\mathbb R^n\) with \(\varphi(0) =0\), \(w\in\mathbb R^m\) are disturbances generated from an unknown exosystem \[ \dot w=Sw,\tag{2} \] and \(C=(1,0,\dots,0)\). This paper deals with global stabilization and disturbance suppression of the class of systems (1-2) using output feedback. The proposed algorithm is able to completely suppress any harmonic disturbances without the knowledge of their amplitudes, phases and the frequencies. The result is global with respect to the state variables and with respect to the unknown frequencies. The only information needed for the exosystem is its order. The proposed control design is based on a new formulation of state estimation. An adaptive internal model is used to generate the contribution of the desired input compensation to a state variable, which is then used in the control design. The control design abandons the parallel structure for output regulation, and it integrates compensation and stabilization components from the first step. This is a direct consequence of using the proposed formulation of state estimation. Another new feature in the design is the introduction of adaptive coefficients in the control design, which allows the removal of the restriction in the unknown frequencies in the disturbances. All the adaptive rates in the system asymptotically approach zero, which implies that the controller will converge to a fixed-parameter type controller. In this work, no uncertainty of the system is considered.