Total least squares (TLS) parameter estimation is an alternative to least squares (LS) estimation when there are errors in both data matrix and observation vector. Especially, when some of the columns, not all, of the data matrix A are free of error, we call it a mixed LS-TLS problem. Accordingly, a sequential algorithm for solving a mixed LS-TLS problem is proposed here. The proposed algorithm employs an efficient algorithm to locate the minimum eigenpair, instead of singular value decomposition (SVD) which is computationally exacting. The proposed algorithm is applied to an accelerometer model to identify error parameters which are very important in inertial navigation systems (INS).


    Zugriff

    Zugriff prüfen

    Verfügbarkeit in meiner Bibliothek prüfen

    Bestellung bei Subito €


    Exportieren, teilen und zitieren



    Titel :

    An LS-TLS-based sequential identification algorithm: Application to an accelerometer


    Beteiligte:
    Chang Wan Jeon, (Autor:in) / Hyoung Joong Kim, (Autor:in) / Jang Gyu Lee, (Autor:in)


    Erscheinungsdatum :

    2001-04-01


    Format / Umfang :

    374219 byte




    Medientyp :

    Aufsatz (Zeitschrift)


    Format :

    Elektronische Ressource


    Sprache :

    Englisch




    MEMS Accelerometer Design Optimization Using Genetic Algorithm

    Krushnasamy, V.S. / Juliet, A.Vimala | Tema Archiv | 2013


    Axle Box Accelerometer Signal Identification and Modelling

    Hoelzl, Cyprien A. / Valencia, Luis David Avendano / Dertimanis, Vasilis K. et al. | Springer Verlag | 2020


    Accelerometer

    Kuttner, Thomas / Rohnen, Armin | Springer Verlag | 2023


    Model Set Identification of Accelerometer with Nonparametric Uncertainty

    Liu, X. / Hu, H. / Huang, X. et al. | British Library Conference Proceedings | 1994