Scribd Logo
Upload

Welcome to Scribd!

  • 95 views

    Lec03 2x2

    Uploaded by

    jacobwatson
    The document discusses robust control theory and robustness to model uncertainties. It introduces key concepts like multiplicative uncertainty, robust stability, robust performance, and the small gain theorem. Robust stability ensures closed-loop stability for all plants in the uncertainty set. Robust performance ensures specifications are met for all uncertain plants. Weights are used to model uncertainty and constrain transfer functions like the sensitivity and complementary sensitivity functions. The small gain theorem and frequency domain specifications can be used to analyze and ensure robust stability and performance.

    Copyright:

    Attribution Non-Commercial (BY-NC)
    Download as PDF, TXT or read online from Scribd

    Lec03 2x2

    Uploaded by

    jacobwatson
    95 views6 pages
    The document discusses robust control theory and robustness to model uncertainties. It introduces key concepts like multiplicative uncertainty, robust stability, robust performance, and the small gain theorem. Robust stability ensures closed-loop stability for all plants in the uncertainty set. Robust performance ensures specifications are met for all uncertain plants. Weights are used to model uncertainty and constrain transfer functions like the sensitivity and complementary sensitivity functions. The small gain theorem and frequency domain specifications can be used to analyze and ensure robust stability and performance.

    Original Description:

    More fantastic control engineering resources - lectures

    Original Title

    lec03-2x2

    Copyright

    © Attribution Non-Commercial (BY-NC)

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    The document discusses robust control theory and robustness to model uncertainties. It introduces key concepts like multiplicative uncertainty, robust stability, robust performance, and the small gain theorem. Robust stability ensures closed-loop stability for all plants in the uncertainty set. Robust performance ensures specifications are met for all uncertain plants. Weights are used to model uncertainty and constrain transfer functions like the sensitivity and complementary sensitivity functions. The small gain theorem and frequency domain specifications can be used to analyze and ensure robust stability and performance.

    Copyright:

    Attribution Non-Commercial (BY-NC)
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    95 views6 pages

    Lec03 2x2

    Uploaded by

    jacobwatson
    The document discusses robust control theory and robustness to model uncertainties. It introduces key concepts like multiplicative uncertainty, robust stability, robust performance, and the small gain theorem. Robust stability ensures closed-loop stability for all plants in the uncertainty set. Robust performance ensures specifications are met for all uncertain plants. Weights are used to model uncertainty and constrain transfer functions like the sensitivity and complementary sensitivity functions. The small gain theorem and frequency domain specifications can be used to analyze and ensure robust stability and performance.

    Copyright:

    Attribution Non-Commercial (BY-NC)
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    of 6

    So far

    wu w

    EL2520 Control Theory and Practice


    Lecture 3: Robustness
    Elling W. Jacobsen School of Electrical Engineering KTH, Stockholm, Sweden

    Fr

    -Fy

    Signal norms, system gains and the small gain theorem The closed-loop system and the design problem Characterized by six transfer functions: need to look at all! Internal stability: stability from all inputs to all outputs (sufficient to check that Fr, S, SG and SFy are all stable) Sensitivity function (suppression of load disturbances) and Complementary sensitivity (noise, robust stability)
    EL2520 Control Theory and Practice Elling W. Jacobsen [email protected]

    EL2520 Control Theory and Practice

    Elling W. Jacobsen [email protected]

    Robustness
    Robustness=Insensitivity to model errors (differences between modelled and actual system behavior)

    Classes of uncertainty
    Parametric uncertainty: model structure known, but some parameters are uncertain
    Dynamic uncertainty: some (often high frequency) dynamics is missing, either by lack of understanding or in order to get a simpler model Often, we have a combination of the two. Convenient to represent in lumped form

    To reason about uncertainty we need to model it! The uncertainty set: defines a family of possible models (quantifies how much we do not know about the system) Would like to establish Robust stability (stability of all plants in uncertainty set) Robust performance (meet specs for all plants in uncertainty set)

    EL2520 Control Theory and Practice

    Elling W. Jacobsen [email protected]

    EL2520 Control Theory and Practice

    Elling W. Jacobsen [email protected]

    Multiplicative uncertainty
    Multiplicative uncertainty

    Robust stability w. multiplicative uncertainty

    Here, I is a family of possible behaviours of the physical plant is any stable transfer function with gain less than one

    Lecture 2: Small-gain theorem interconnection stable if wIT internally stable, stable, and Robust stability: closed-loop stability for all G2 I
    EL2520 Control Theory and Practice Elling W. Jacobsen [email protected]

    Satisfied if
    EL2520 Control Theory and Practice Elling W. Jacobsen [email protected]

    Example: uncertain gain


    Consider the set of possible plants

    Example: uncertain zero location


    Consider the set of possible plants

    Can be put into standard form via

    Can re-write as

    where

    Note: here it is enough to let be real (in standard form is complex)

    Note: wI is now frequency dependent, still real

    EL2520 Control Theory and Practice

    Elling W. Jacobsen [email protected]

    EL2520 Control Theory and Practice

    Elling W. Jacobsen [email protected]

    Alternative approach to obtain weight


    Note that multiplicative uncertainty class

    Example
    Consider the uncertain system

    can be re-written as

    with nominal plant

    Sample uncertainties (dotted) and corresponding wI (dashed) Thus, the uncertainty about the system captured by WI if

    (must first pick a nominal model G)

    EL2520 Control Theory and Practice

    Elling W. Jacobsen [email protected]

    EL2520 Control Theory and Practice

    Elling W. Jacobsen [email protected]

    Example: robust stability


    Consider the following nominal plant and controller

    Example: robust stability

    Relative error and assume that one extreme possible plant is

    is around 0.33 for low frequencies and 5.25 at high frequencies. Suggests weight

    EL2520 Control Theory and Practice

    Elling W. Jacobsen [email protected]

    EL2520 Control Theory and Practice

    Elling W. Jacobsen [email protected]

    Example: robust stability


    Uncertainty weight wI and complementary sensitivities for two sets of controller parameters

    Robust stability in the Nyquist curve


    Uncertain system: G(i) takes one of several possible values at each frequency a family of Nyquist curves

    First setting (T1) is not robustly stable, second setting (T2) is. Robust stability if uncertainty regions do not encircle -1 point

    EL2520 Control Theory and Practice

    Elling W. Jacobsen [email protected]

    EL2520 Control Theory and Practice

    Elling W. Jacobsen [email protected]

    Complementary sensitivity in Nyquist


    Constraint on complementary sensitivity

    Frequency domain specifications


    Forbidden area

    also yields circles that should be avoided by the Nyquist curve.


    Forbidden area

    Circles centered at
    EL2520 Control Theory and Practice

    with radius
    Elling W. Jacobsen [email protected]

    Can we choose weights wS, wT (forbidden areas) freely? No, there are many constraints and limitations!
    EL2520 Control Theory and Practice Elling W. Jacobsen [email protected]

    Extension: shaping the gang of six


    Can shape all relevant transfer functions (in the gang of six)

    Robust performance
    Nominal performance specified in terms of sensitivity function

    Robust performance

    Since

    This is the topic of Computer Exercise 2!


    Worst-case is such that 1+L and wI L point in opposite directions

    EL2520 Control Theory and Practice

    Elling W. Jacobsen [email protected]

    EL2520 Control Theory and Practice

    Elling W. Jacobsen [email protected]

    Robust performance contd


    Robust performance

    Robust stability and performance


    In summary

    Can be expressed as
    Note that nominal performance and robust stability implies Sometimes approximated by the mixed sensitivity constraint Thus, robust stability and nominal performance implies close to robust performance. Note! This only holds in SISO case.

    EL2520 Control Theory and Practice

    Elling W. Jacobsen [email protected]

    EL2520 Control Theory and Practice

    Elling W. Jacobsen [email protected]

    Summary
    Robustness insensitivity to model errors
    Can guarantee robustness if we model (or bound) uncertainty general tool: small gain theorem sometimes need to pull out uncertainty by hand sometimes, can fall back onto standard forms (e.g. multiplicative input uncertainty) Robustness typically introduces new constraints on T Robust performance: acceptable S, despite uncertainties.

    EL2520 Control Theory and Practice

    Elling W. Jacobsen [email protected]

    You might also like