scientific article; zbMATH DE number 2059913

From MaRDI portal
Publication:4458200





Consider a class of nonlinear systems that can be transformed into \[ \dot x=A_cx + b(\theta)\sigma(y)u+\sum_{i=1}^p\psi_i(y)a_i+\psi_0(y),\quad y(t)= C_cx, \] where \(\sigma(\cdot)\in C(\mathbb R,\mathbb R)\), \(\psi_i(\cdot)\in C(\mathbb R,\mathbb R^n)\) and \(A_c\) is the Jordan matrix in \(\mathbb R^{n\times n}\), \(b(\theta)=(0,\dots, 0,b_m,\dots,b_0)^T\in\mathbb R^n\), \(C_c=(1,0,\dots,0)^T\in\mathbb R^n\), with system relative degree \(\rho = n - m\). The control objective is to force the system output to asymptotically track a given reference signal \(y_r(t)\) in the presence of unknown \(a_i\) and \(b_j\), \((i = 1,\dots,p\), \(j = 0,\dots,m)\). To solve the problem, a scheme using the Nussbaum gain in the backstepping design is proposed. It is shown that the controller obtained by the proposed design scheme can make the output of the system asymptotically track a given signal while all the signals in the whole adaptive control system remain stable. The stability analysis here is obtained without imposing any growth conditions on system nonlinearities. Another feature of the proposed scheme is the way of constructing the Nussbaum gain. Unlike the conventional approaches as in all previous results using Nussbaum gains, an augmented tracking error is no longer used here and the dynamics about the Nussbaum gain variable is determined only by a first-order system. This ensures that the order of the auxiliary system, that is used to construct the Nussbaum gain, is reduced to the minimum.











This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4458200)