Robust adaptive tracking control of uncertain discrete time systems

From MaRDI portal





The author considers the class of uncertain SISO discrete-time systems described by \[ (\hat a(z) -\Delta)y(t) = z^d{\hat b}(z)u(t) + \nu(t), \tag{1} \] where \(y(t)\), \(u(t)\) and \(\nu(t)\) denote the output, input and external disturbance, respectively, \(t\) is the discrete time, \(d\) is the I/O delay which takes integer values greater than 1, \(\Delta\) denotes the unmodelled dynamics, \[ \hat a(z) = 1 + a(1)z^1 +\dots + a(n)z^n,\quad \hat b(z) = b(0) + b(1)z^1 +\dots + b(m)z^m. \] It is assumed that the structural parameters \(n\), \(m\), \(d\) are known a priori. The polynomials \(\hat a(z)\) and \(\hat b(z)\) do not have the same unstable zeros; the unmodelled dynamics \(\Delta\in D_{\infty,FM}(w_\Delta)\), where \(D_{\infty,FM}(w_\Delta)=\left\{\Delta:\Delta\right.\) is the casual time-varying operator with finite-memory, and satisfies \( \left. \|\Delta\|_1 = \sup_{x\in\ell_\infty}(\|\Delta x\|_\infty/\| x\|_\infty)\leqslant w_\Delta\right\}\), which implies that there exists \(\mu > 0\) such that \(| \Delta x(t)| \leqslant w_\Delta\{\max_{0\leqslant k\leqslant\mu}| x(t-k)| \}\), where \(\Delta x(t) =\sum_{k=t-\mu}\Delta(t - k) x(k)\). The external disturbances satisfy \(\sup_t| \nu(t)| \leqslant w_\nu\). For the system (1), an indirect adaptive tracking controller is obtained by frozen-time controllers that at each time optimally robustly stabilize the estimated models of the plant and minimize the worst-case steady-state absolute value of the tra\-c\-king error of the estimated model over the model uncertainty. Based on \(\ell_\infty\) to \(\ell_\infty\) stability and performance of slowly varying system found in the literature, the proposed adaptive tracking scheme is shown to have good robust stability. Moreover, a computable upper bound on the size of the unstructured uncertainty permitted by the adaptive system and a computable tight upper bound on asymptotic robust steady-state tracking performance are provided. A numerical example to illustrate the adaptive control scheme and verify the theoretical results is presented.




Cited in
(24)








This page was built for publication: Robust adaptive tracking control of uncertain discrete time systems

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5704787)