A Time-Frequency Local Polynomial Approach to FRM Estimation from Incomplete Data

N.J. Dirkx, Koen Tiels, Tom Oomen

Research output: Contribution to journalConference articlepeer-review

83 Downloads (Pure)


Frequency Response Matrix (FRM) estimation from measured data is an important step towards the control of complex systems, including motion and thermal systems. Missing samples in the measured data records, e.g., due to sensor failure or faulty data transmission, often occur. In this paper, a method is presented for the nonparametric FRM identification of multiple-inputs multiple-outputs (MIMO) systems from incomplete and noisy data records. The method exploits time- and frequency-domain localizing wavelets to accurately estimate the FRM and its covariance from the time-frequency plane. Good performance is demonstrated in a simulation study.
Original languageEnglish
Pages (from-to)3942-3947
Number of pages6
Issue number2
Publication statusPublished - 1 Jul 2023
Event22nd World Congress of the International Federation of Automatic Control (IFAC 2023 World Congress) - Yokohama, Japan
Duration: 9 Jul 202314 Jul 2023
Conference number: 22


This work was supported by the Research Programme VIDI under Project 15698, partly financed by the NWO.

FundersFunder number
Nederlandse Organisatie voor Wetenschappelijk Onderzoek


    • Frequency response function identification
    • linear systems
    • missing data
    • multiple-inputs multiple-outputs
    • transient estimation


    Dive into the research topics of 'A Time-Frequency Local Polynomial Approach to FRM Estimation from Incomplete Data'. Together they form a unique fingerprint.

    Cite this