A Newton-Raphson algorithm for calculating the effects of changes in weights on an H design

Alexander Lanzon*, Brian D.O. Anderson, Xavier Bombois

*Corresponding author for this work

    Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

    Abstract

    When weights are adjusted in an H design, corresponding modifications occur in the synthesised H controller and the resulting closed-loop transfer function matrices of interest. This article proposes an algorithm that allow us to compute the corresponding modifications hi the controller and closed-loop transfer functions via calculations that are simpler/faster than solving two brand new Riccati equations from scratch. This is particularly beneficial (in terms of both speed and numerical reliability of the solutions) for high order systems and in iterative adaptive control algorithms that involve weight adjustments at each sampling interval. The algorithm proposed here builds upon our previous results presented in [1] and now allow us to also handle medium-size and large weight adjustments, as opposed to only small weight adjustments.

    Original languageEnglish
    Title of host publicationProceedings of the 44th IEEE Conference on Decision and Control, and the European Control Conference, CDC-ECC '05
    Pages3657-3662
    Number of pages6
    DOIs
    Publication statusPublished - 2005
    Event44th IEEE Conference on Decision and Control, and the European Control Conference, CDC-ECC '05 - Seville, Spain
    Duration: 12 Dec 200515 Dec 2005

    Publication series

    NameProceedings of the 44th IEEE Conference on Decision and Control, and the European Control Conference, CDC-ECC '05
    Volume2005

    Conference

    Conference44th IEEE Conference on Decision and Control, and the European Control Conference, CDC-ECC '05
    Country/TerritorySpain
    CitySeville
    Period12/12/0515/12/05

    Fingerprint

    Dive into the research topics of 'A Newton-Raphson algorithm for calculating the effects of changes in weights on an H design'. Together they form a unique fingerprint.

    Cite this