Fernando H. D. Guaracy

This article presents a set of sufficient conditions that allows synchronization protocols for discrete-time homogeneous linear multi-agent systems based on state feedback to be not only implemented but also designed in a fully distributed manner, that is, without the need to use global information about the entire communication topology. The results are obtained from a guaranteed synchronization region determined from an auxiliary problem that considers only the dynamics of a single agent. Numerical examples are presented to illustrate the application of the proposed conditions.

This article presents an analysis of mean-square stability and robustness margins for networked systems. We consider an architecture in which the controller-to-actuator channel is subject to packet losses, and a zero-input strategy is used to deal with the losses. For this situation, we obtain guaranteed mean-square gain and phase margins based on the $\mathcal {H}_{\infty }$ norm of sensitivity and complementary sensitivity functions. We also analyze guaranteed mean-square $S$-based and $T$-based disk margins, which in the case of single-input, single-output systems and losses modelled by Bernoulli processes can be translated into a tool based on Nyquist plots. A numerical example is presented to illustrate the use of the proposed results.

This article concerns the problem of synchronization of homogeneous linear multi-agent systems considering a robust control and small gain theorem perspective. The analyses are performed based on an auxiliary state feedback problem in which only the dynamics of a single agent is considered. This point of view allows an unified approach to continuous- and discrete-time cases and provides insights that, in addition to enabling the analysis of feasibility of the synchronization problem in terms of a guaranteed synchronization region, assists in choosing the distributed protocol design parameters. Numerical examples are presented to illustrate the analyses performed.

In this article, we presented guaranteed independent and uniform multivariable stability margins for the $\mathcal {H}_{\infty }$ /LTR control, both in continuous and discrete time. In previous works, it was shown that the $\mathcal {H}_{\infty }$ /LTR control generalizes the linear quadratic Gaussian (LQG)/LTR setup, allowing the designer to better impose bounds on the complementary sensitivity frequency peak in the loop-shaping procedure. In this article, we extend this result by showing that $\mathcal {H}_{\infty }$ /LTR control is able to recover better robustness properties than those recovered in the LQG/LTR case. Numerical examples are presented for multivariable and monovariable cases.

In this paper a discrete-time static H _∞ loop shaping control method for linear parameter-varying (LPV) systems is presented. The formulation aims to find a static parameter dependent stabilizing controller which yields guaranteed performance. Using quadratic H _∞ performance, a set of sufficient conditions is provided in a two-step LMI (Linear Matrix Inequality) approach. The first step consists in an extension of the left coprime factorization for discrete-time LPV systems, in which a particular H _∞ filtering problem is solved. In the second step, sufficient conditions for the solution of the static H _∞ loop shaping control problem are provided. Besides considering parameter dependency, these conditions also include nonparametric uncertainties, resulting in more robust closed-loop systems. The effectiveness of the method is illustrated with a numerical example.

This paper presents a new approach on the plant augmentation and target feedback loop parameterization for the discrete-time LQG/LTR control. With the proposed procedure, the bound on the sensitivity of the control loop can be associated with the behavior of the inverse of any real stable pole, ensuring good properties of disturbance rejection and tracking of reference signals in situations where the expansion of the plant by integrators is prohibitive or undesirable.