Optimal adaptive controller for systems with delay

From MaRDI portal





The authors of this paper consider optimal control of the system \[ \begin{cases} A(z)y_n=z^dB(z)u_n+w_n\quad & \text{for }n>0\\ u_n=y_n=w_n=0 & \text{for }n\leq 0,\end{cases} \] where \(d\) is system delay, \(A(z)\) and \(B(z)\) are polynomials, \(y_n\), \(u_n\) and \(w_n\) are system output, input and random disturbance respectively. A method for \(d\)-step-ahead stochastic control is developed using input matching and a least-squares estimation algorithm. The authors' main results are the following: (1) The identification convergence rate of parameter estimation is determined for the \(d\)-step-ahead stochastic system: \[ \| \theta-\theta_{N+d}\|^2=O\left(\frac{\log n}{N}\right) \] almost surely, where \(\theta\) is the parameter, and \(\theta_{N+d}\) is the \(N^{\text{th}}\) estimate; (2) Without any assumption of persistent excitation, the \(d\)-step-ahead closed-loop system is globally stable; (3) The adaptive controller converges to the \(d\)-step-ahead optimal controller; and (4) Consistency of the parameter identification based on the \(d\)-step-ahead least-squares algorithm is established.











This page was built for publication: Optimal adaptive controller for systems with delay

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4667783)