Non-Parametric Estimation of Correlation Functions

Rune Brincker, Anders Rytter, Steen Krenk

    Research output: Book/ReportReportResearchpeer-review

    120 Downloads (Pure)

    Abstract

    In this paper three methods of non-parametric correlation function estimation are reviewed and evaluated: the direct method, estimation by the Fast Fourier Transform and finally estimation by the Random Decrement technique. The basic ideas of the techniques are reviewed, sources of bias are pointed out, and methods to prevent bias are presented. The techniques are evaluated by comparing their speed and accuracy on the simple case of estimating auto-correlation functions for the response of a single degree-of-freedom system loaded with white noise.
    Original languageEnglish
    Place of PublicationAalborg
    PublisherDept. of Building Technology and Structural Engineering, Aalborg University
    Number of pages15
    Publication statusPublished - 1991
    SeriesFracture and Dynamics
    Number31
    VolumeR9120
    ISSN0902-7513

    Fingerprint

    White noise
    Autocorrelation
    Fast Fourier transforms

    Bibliographical note

    Submitted to Experimental Mechanics

    PDF for print: 20 pp.

    Keywords

    • Fast Fourier Transform
    • RDD

    Cite this

    Brincker, R., Rytter, A., & Krenk, S. (1991). Non-Parametric Estimation of Correlation Functions. Aalborg: Dept. of Building Technology and Structural Engineering, Aalborg University. Fracture and Dynamics, No. 31, Vol.. R9120
    Brincker, Rune ; Rytter, Anders ; Krenk, Steen. / Non-Parametric Estimation of Correlation Functions. Aalborg : Dept. of Building Technology and Structural Engineering, Aalborg University, 1991. 15 p. (Fracture and Dynamics; No. 31, Vol. R9120).
    @book{e55efbb07cbe11dc97b9000ea68e967b,
    title = "Non-Parametric Estimation of Correlation Functions",
    abstract = "In this paper three methods of non-parametric correlation function estimation are reviewed and evaluated: the direct method, estimation by the Fast Fourier Transform and finally estimation by the Random Decrement technique. The basic ideas of the techniques are reviewed, sources of bias are pointed out, and methods to prevent bias are presented. The techniques are evaluated by comparing their speed and accuracy on the simple case of estimating auto-correlation functions for the response of a single degree-of-freedom system loaded with white noise.",
    keywords = "RDD, Fast Fourier Transform, Fast Fourier Transform, RDD",
    author = "Rune Brincker and Anders Rytter and Steen Krenk",
    note = "Submitted to Experimental Mechanics PDF for print: 20 pp.",
    year = "1991",
    language = "English",
    series = "Fracture and Dynamics",
    publisher = "Dept. of Building Technology and Structural Engineering, Aalborg University",
    number = "31",
    address = "Denmark",

    }

    Brincker, R, Rytter, A & Krenk, S 1991, Non-Parametric Estimation of Correlation Functions. Fracture and Dynamics, no. 31, vol. R9120, Dept. of Building Technology and Structural Engineering, Aalborg University, Aalborg.

    Non-Parametric Estimation of Correlation Functions. / Brincker, Rune; Rytter, Anders; Krenk, Steen.

    Aalborg : Dept. of Building Technology and Structural Engineering, Aalborg University, 1991. 15 p. (Fracture and Dynamics; No. 31, Vol. R9120).

    Research output: Book/ReportReportResearchpeer-review

    TY - RPRT

    T1 - Non-Parametric Estimation of Correlation Functions

    AU - Brincker, Rune

    AU - Rytter, Anders

    AU - Krenk, Steen

    N1 - Submitted to Experimental Mechanics PDF for print: 20 pp.

    PY - 1991

    Y1 - 1991

    N2 - In this paper three methods of non-parametric correlation function estimation are reviewed and evaluated: the direct method, estimation by the Fast Fourier Transform and finally estimation by the Random Decrement technique. The basic ideas of the techniques are reviewed, sources of bias are pointed out, and methods to prevent bias are presented. The techniques are evaluated by comparing their speed and accuracy on the simple case of estimating auto-correlation functions for the response of a single degree-of-freedom system loaded with white noise.

    AB - In this paper three methods of non-parametric correlation function estimation are reviewed and evaluated: the direct method, estimation by the Fast Fourier Transform and finally estimation by the Random Decrement technique. The basic ideas of the techniques are reviewed, sources of bias are pointed out, and methods to prevent bias are presented. The techniques are evaluated by comparing their speed and accuracy on the simple case of estimating auto-correlation functions for the response of a single degree-of-freedom system loaded with white noise.

    KW - RDD

    KW - Fast Fourier Transform

    KW - Fast Fourier Transform

    KW - RDD

    M3 - Report

    T3 - Fracture and Dynamics

    BT - Non-Parametric Estimation of Correlation Functions

    PB - Dept. of Building Technology and Structural Engineering, Aalborg University

    CY - Aalborg

    ER -

    Brincker R, Rytter A, Krenk S. Non-Parametric Estimation of Correlation Functions. Aalborg: Dept. of Building Technology and Structural Engineering, Aalborg University, 1991. 15 p. (Fracture and Dynamics; No. 31, Vol. R9120).