An LMI approach to \(\mathcal D\)-stability and \(\mathcal D\)-stabilization of linear discrete singular systems with state delay (Q654634)

Consider a linear discrete singular time-delay system of the form \[ Ex(k+1)=Ax(k)+A_{d}x(k-d)+Bu(k), \] where \(x\in \mathbb R^{n}\) and \(u\in \mathbb R^{m}\) denote the state vector and the control input, respectively; \(E\in \mathbb R^{n\times n}\) may be singular with rank \(\leq n,\) \(A\),\(A_{d}\) and \(B\) are real matrices, and \(d\) is a known positive integer time delay of the system. Its characteristic polynomial without input is given by \[ F(z)=\det \left( zE-A-A_{d}z^{-d}\right) . \] This system is said to be \(D(\alpha ,r)\)-stable if all the finite roots of \( F(z)\) lie inside the disk \(D(\alpha ,r)\) with center \(\alpha \) and radius \(r\) where \(\left| \alpha \right| +r<1,\) and is said to be \(D(\alpha ,r)\) stabilizable if there exists a matrix \(K\in \mathbb R^{m\times n}\) such that the closed loop system \[ Ex(k+1)=(A+BK)x(k)+A_{d}x(k-d) \] is \(D(\alpha ,r)\)-stable, where \(u(k)=Kx(k)\). By using the state augmentation technique, a necessary and sufficient \( D(\alpha ,r)\)-stability condition and a sufficient \(D(\alpha ,r)\)-stabilization condition are proposed. Based on these conditions, `quadratic \(D(a,r)\)-stability and \(D(\alpha ,r)\)-stabilization conditions' for linear discrete singular time-delay systems with norm-bounded uncertainties and interval uncertainties are obtained, respectively. Both the stability and stabilization conditions are proposed in terms of linear matrix inequalities (LMI). Examples are also given for illustrations.