A novel distributed algorithm for estimation and control of large-scale systems

In this paper we propose a novel algorithm based on linear matrix inequalities for the design of distributed controllers and state estimators for large-scale systems inspired by linear quadratic regulators and Kalman filters, respectively. With respect to similar state-of-the art methods, the scheme proposed here allows to reduce the conservativeness due to the approximations used for the covariance distributed iterative computation. The theoretical properties of the proposed scheme are thoroughly investigated. The controllers and observers obtained using the proposed approach are applied to a simulated dynamical system and their performances are thoroughly compared to those obtained with state-of-the-art schemes, showing the potentialities of the scheme.