This paper discusses a class of large-scale systems composed of several similar subsystems interconnected with an external system in a symmetrical fashion. A method of determining the set of fixed modes of such a system with respect to decentralized (or partly decentralized) control is presented in terms of the modified subsystems. The completeness of these systems is studied, and a simple condition for such a system to be complete is derived.

This paper is concerned with the problem of suboptimal stable mixed H2/H∞ control for linear time-invariant systems. The designed controller with the same order as that of the considered system is required to satisfy a prescribed H∞ performance bound or a prescribed degree of stability. By reducing the stable controller synthesis problem to a multiobjective state feedback control problem for two different state models, sufficient conditions for the solvability of the considered problem are given in terms of solutions to algebraic Riccati equations and matrix inequalities. An LMI-based iterative algorithm is developed to solve the stable controller synthesis problem. The proposed algorithm is shown to be convergent. Examples are given to illustrate the proposed methods.

This paper is concerned with the single-contingency reliable controller design problem for nonlinear systems. A procedure for designing reliable nonlinear controllers is presented for nonlinear systems with strictly redundant sensors and strictly redundant actuators. The resulting closed-loop nonlinear systems are reliable in the sense that they provide guaranteed local internal stability and H∞ performance not only when all sensors and actuators are operational but also when any one, but only one, sensor or actuator experiences an outage. An example is given to illustrate the design procedure and the effectiveness of the proposed methods.

This note is concerned with the problem of a robust non-fragile Kalman filter design for a class of uncertain linear systems with norm-bounded uncertainties. The designed state estimator can tolerate multiplicative uncertainties in the state estimator gain matrix. The robust nonfragile state estimator designs are given in terms of solutions to algebraic Riccati equations. The designs guarantee known upper bounds on the steady-state error covariance. A numerical example is given to illustrate the results.