New dilated LMI characterizations for continuous-time multiobjective controller synthesis (Q1765190)

The authors' main objective is to extend to the continuous-time case a linear matrix inequality (LMI) stabilizability condition originally available in the discrete-time case only. The main feature of the LMI condition is a decoupling between Lyapunov variables (ensuring stability or performance) and controller variables. The extension proposed in this paper is only partial and incomplete, but covers a non-trivial multiobjective design problem mixing \(H_2\) performance and pole location in a given region. It is shown that this design problem can be formulated as a nonlinear matrix inequality. This matrix inequality is non-convex in general, but it becomes a convex LMI problem when fixing a scalar parameter. Interestingly, it is shown in Theorem 6 that this LMI is always less conservative than previously available LMI design conditions, irrespective of the choice of the scalar parameter. This is the main contribution of this technical communique.