Electrical and Electronic Engineering - Research Publications

Permanent URI for this collection

http://hdl.handle.net/11343/362

Filters
Reset filters

Settings
Statistics
Citations
Author: Nesic, Dragan × Start date: 2015 × Type: Journal Article ×

Search Results

Now showing 1 - 10 of 67
  • Item
    No Preview Available
    Decentralized event-triggered estimation of nonlinear systems
    Petri, E ; Postoyan, R ; Astolfi, D ; Nešić, D ; Heemels, WPMH (Elsevier BV, 2024-02-01)
  • Item
    No Preview Available
    Stability analysis of optimal control problems with time-dependent costs?
    Benahmed, S ; Postoyan, R ; Granzotto, M ; Busoniu, L ; Daafouz, J ; Nesic, D (PERGAMON-ELSEVIER SCIENCE LTD, 2023-11)
  • Item
    No Preview Available
    TRACKING AND REGRET BOUNDS FOR ONLINE ZEROTH-ORDER EUCLIDEAN AND RIEMANNIAN OPTIMIZATION
    Maass, A ; Manzie, C ; Nesic, D ; Manton, JH ; Shames, I (SIAM PUBLICATIONS, 2022)
  • Item
    No Preview Available
    Event-Triggered Control Through the Eyes of a Hybrid Small-Gain Theorem
    Maass, AI ; Wang, W ; Nesic, D ; Postoyan, R ; Heemels, M (IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC, 2023-10)
  • Item
    No Preview Available
    Learning-Based Adaptive Control for Stochastic Linear Systems With Input Constraints
    Siriya, S ; Zhu, J ; Nesic, D ; Pu, Y (IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC, 2023)
  • Item
    No Preview Available
    A Multi-Processor Implementation for Networked Control Systems
    Maass, AI ; Wang, W ; Nesic, D ; Tan, Y ; Postoyan, R (IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC, 2023)
  • Item
    No Preview Available
    On state estimation for nonlinear systems under random access wireless protocols
    Maass, AI ; Nesic, D ; Postoyan, R ; Tan, Y (SPRINGER LONDON LTD, 2023-03-01)
    This article is dedicated to Eduardo D. Sontag on the occasion of his 70th birthday. We build upon fundamental stability concepts developed by Sontag, such as input-to-state stability and its related properties, to study a relevant application in industrial internet of things, namely estimation for wireless networked control systems. Particularly, we study emulation-based state estimation for nonlinear plants that communicate with a remote observer over a shared wireless network subject to packet losses. To reduce bandwidth usage, a stochastic communication protocol is employed to determine which node should be given access to the network. Each node has a different successful transmission probability. We describe the overall closed-loop system as a stochastic hybrid model, which allows us to capture the behaviour both between and at transmission instants, whilst covering network features such as random transmission instants, packet losses and stochastic scheduling. We then provide sufficient conditions on the transmission rate that guarantee an input-to-state stability property (in expectation) for the corresponding estimation error system. We illustrate our results in the design of circle criterion observers.
  • Item
    Thumbnail Image
    Stability and Persistent Excitation in Signal Sets
    Lee, T-C ; Tan, Y ; Nesic, D (IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC, 2015-05)
    Persistent excitation (PE) conditions have been widely used to analyze stability properties of various parameter identification algorithms and to establish uniform global asymptotic stability (UGAS) for a large class of nonlinear time-varying systems. In order to generalize such conditions to a more general setting, a new PE condition is proposed with three basic ingredients: a signal set to represent a family of time functions (e.g., trajectories); a pseudo distance measure to describe the convergence; and some binary relations (e.g., state-to-output relations). Closely related to detectability, this PE condition is a necessary condition to guarantee UGAS. Under uniform global stability and an integral inequality, it becomes a sufficient condition of UGAS. A novel concept: M-pair, which aims at simplifying the checking of the PE condition, is introduced. By using M-pair, it is possible to simplify the structure of the referred signal set (in the spirit of the classic Krasovskii-LaSalle theorem) and to extend the dimension of the reference signal set (similar to the Matrosov theorem). Thus, the framework of M-pair not only unifies these well-known results, but also generates more flexibility in checking the PE conditions. When applied to nonlinear switched systems, three new tools to verify the PE condition are obtained. Finally, an example illustrates that a nonlinear time-varying switched system with arbitrary switching can be shown to be UGAS without using a common Lyapunov function.
  • Item
    Thumbnail Image
    Event-triggered tracking control of unicycle mobile robots
    Postoyan, R ; Bragagnolo, MC ; Galbrun, E ; Daafouz, J ; Nesic, D ; Castelan, EB (PERGAMON-ELSEVIER SCIENCE LTD, 2015-02)
    We investigate the stabilization of time-varying trajectories for unicycle mobile robots using event-triggered controllers. We follow an emulation-like approach in the sense that we first synthesize the controller while ignoring the communication constraints and we then derive an appropriate triggering condition. The solutions to the robot model are proved to practically converge towards the given reference trajectory, under some condition on the latter. Furthermore, the existence of a uniform minimum amount of time between any two transmissions is ensured. Afterward, experimental results are presented where the controller has been implemented on a remote computer which transmits its output to the mobile robot via an IEEE 802.11g wireless network. The proposed event-triggering strategy is able to significantly reduce the need for communication compared to a classical time-triggered setup while ensuring similar, if not better, tracking performances.
  • Item
    Thumbnail Image
    A Framework for the Event-Triggered Stabilization of Nonlinear Systems
    Postoyan, R ; Tabuada, P ; Nesic, D ; Anta, A (IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC, 2015-04)
    Event-triggered control consists of closing the feedback loop whenever a predefined state-dependent criterion is satisfied. This paradigm is especially well suited for embedded systems and networked control systems since it is able to reduce the amount of communication and computation resources needed for control, compared to the traditional periodic implementation. In this paper, we propose a framework for the event-triggered stabilization of nonlinear systems using hybrid systems tools, that is general enough to encompass most of the existing event-triggered control techniques, which we revisit and generalize. We also derive two new event-triggering conditions which may further enlarge the inter-event times compared to the available policies in the literature as illustrated by two physical examples. These novel techniques exemplify the relevance of introducing additional variables for the design of the triggering law. The proposed approach as well as the new event-triggering strategies are flexible and we believe that they can be used to address other event-based control problems.