Reference model decomposition in direct adaptive control

H. Butler, G. Honderd, J. van Amerongen

    Research output: Contribution to journalArticleAcademicpeer-review

    9 Citations (Scopus)
    71 Downloads (Pure)

    Abstract

    This paper introduces the method of reference model decomposition as a way to improve the robustness of model reference adaptive control systems (MRACs) with respect to unmodelled dynamics with a known structure. Such unmodelled dynamics occur when some of the nominal plant dynamics are purposely neglected in the controller design with the aim of keeping the controller order low. One of the effects of such undermodelling of the controller is a violation of the perfect model-matching condition of the primary controller. The decomposition can be seen as a way to adjust the reference model output (and hence the control goal) to the actual model-matching capabilities. It is shown that the decomposition alleviates the negative effects unmodelled dynamics have on the error equation. Simulation examples illustrate the decomposition design steps and show the obtained improvements.
    Original languageUndefined
    Pages (from-to)199-217
    Number of pages18
    JournalInternational journal of adaptive control and signal processing
    Volume5
    Issue number3
    DOIs
    Publication statusPublished - 1991

    Keywords

    • METIS-111822
    • IR-70933
    • Robustness
    • Model reference adaptive control
    • Reference model adjustment
    • Unmodelled dynamics

    Cite this