Force control for flexible robots using neural networks

Joseph Borowiec, Antonios Tzes

    Research output: Contribution to journalConference article

    Abstract

    Force control for flexible link robots using neural networks is considered in this article. The nonlinear dynamics of the robot manipulator are identified through a recurrent neural network (RNN), which is trained in an off-line manner. Inversion of the RNN-based model dynamics leads to a feedforward component. The feedback controller gains are derived from the minimization of a discrete linear quadratic cost functional, subject to the model dynamics inferred by the linearization of the neural network along the desired trajectory. Sufficient conditions for temporal gain switching bounds are provided. The proposed control scheme is employed in simulation studies on a two link rigid-flexible manipulator.

    Original languageEnglish (US)
    Pages (from-to)1950-1954
    Number of pages5
    JournalProceedings of the American Control Conference
    Volume3
    StatePublished - Dec 1 1999
    EventProceedings of the 1999 American Control Conference (99ACC) - San Diego, CA, USA
    Duration: Jun 2 1999Jun 4 1999

    Fingerprint

    Recurrent neural networks
    Force control
    Dynamic models
    Robots
    Neural networks
    Flexible manipulators
    Linearization
    Manipulators
    Trajectories
    Feedback
    Controllers

    ASJC Scopus subject areas

    • Electrical and Electronic Engineering

    Cite this

    Force control for flexible robots using neural networks. / Borowiec, Joseph; Tzes, Antonios.

    In: Proceedings of the American Control Conference, Vol. 3, 01.12.1999, p. 1950-1954.

    Research output: Contribution to journalConference article

    Borowiec, Joseph ; Tzes, Antonios. / Force control for flexible robots using neural networks. In: Proceedings of the American Control Conference. 1999 ; Vol. 3. pp. 1950-1954.
    @article{02f98819a1fc42378449d10cfd66cbcc,
    title = "Force control for flexible robots using neural networks",
    abstract = "Force control for flexible link robots using neural networks is considered in this article. The nonlinear dynamics of the robot manipulator are identified through a recurrent neural network (RNN), which is trained in an off-line manner. Inversion of the RNN-based model dynamics leads to a feedforward component. The feedback controller gains are derived from the minimization of a discrete linear quadratic cost functional, subject to the model dynamics inferred by the linearization of the neural network along the desired trajectory. Sufficient conditions for temporal gain switching bounds are provided. The proposed control scheme is employed in simulation studies on a two link rigid-flexible manipulator.",
    author = "Joseph Borowiec and Antonios Tzes",
    year = "1999",
    month = "12",
    day = "1",
    language = "English (US)",
    volume = "3",
    pages = "1950--1954",
    journal = "Proceedings of the American Control Conference",
    issn = "0743-1619",
    publisher = "Institute of Electrical and Electronics Engineers Inc.",

    }

    TY - JOUR

    T1 - Force control for flexible robots using neural networks

    AU - Borowiec, Joseph

    AU - Tzes, Antonios

    PY - 1999/12/1

    Y1 - 1999/12/1

    N2 - Force control for flexible link robots using neural networks is considered in this article. The nonlinear dynamics of the robot manipulator are identified through a recurrent neural network (RNN), which is trained in an off-line manner. Inversion of the RNN-based model dynamics leads to a feedforward component. The feedback controller gains are derived from the minimization of a discrete linear quadratic cost functional, subject to the model dynamics inferred by the linearization of the neural network along the desired trajectory. Sufficient conditions for temporal gain switching bounds are provided. The proposed control scheme is employed in simulation studies on a two link rigid-flexible manipulator.

    AB - Force control for flexible link robots using neural networks is considered in this article. The nonlinear dynamics of the robot manipulator are identified through a recurrent neural network (RNN), which is trained in an off-line manner. Inversion of the RNN-based model dynamics leads to a feedforward component. The feedback controller gains are derived from the minimization of a discrete linear quadratic cost functional, subject to the model dynamics inferred by the linearization of the neural network along the desired trajectory. Sufficient conditions for temporal gain switching bounds are provided. The proposed control scheme is employed in simulation studies on a two link rigid-flexible manipulator.

    UR - http://www.scopus.com/inward/record.url?scp=0033284412&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=0033284412&partnerID=8YFLogxK

    M3 - Conference article

    VL - 3

    SP - 1950

    EP - 1954

    JO - Proceedings of the American Control Conference

    JF - Proceedings of the American Control Conference

    SN - 0743-1619

    ER -