Robust LQ regulator for jump linear systems with uncertain parameters (Q1290884)

The authors consider the problem of robust stabilizability of a class of linear systems with Markovian jumping parameter and unknown but bounded uncertainties described by \[ \dot x(t)= [A(r(t))+\Delta A(r(t), a,t)]x(t)+ [B(r(t))+\Delta B(r(t), a,t)]u(t),\tag{1} \] where \(x\in\mathbb{R}^n\), and \(u\in\mathbb{R}^r\) are state and control vectors, respectively; \(A(r(t)),\Delta A(r(t), a,t)\), \(B(r(t))\) and \(\Delta B(r(t), a,t)\) are matrices of appropriate dimensions; \(r(t)\) represents a homogeneous continuous-time discrete-state Markov process taking values in a finite set \({\mathcal B}= \{1,2,..., s\}\). For each \(r(t)\in{\mathcal B}\), \(\Delta A(r(t),a,t)\) and \(\Delta B(r(t), a,t)\) represent the system's uncertainties. The vector parameter \(a\) lies within a prespecified bounded and connected set \({\mathfrak C}\in R^{qa}\). The authors establish a condition that the uncertainties should satisfy in order that the closed-loop of the uncertain system with a feedback control law obtained from the LQ control problem using the nominal model remains stable. They do the same analysis when the uncertainties are of structured type.