Optimal sensor placement of triaxial accelerometers for modal expansion

Vesa Nieminen (Corresponding Author), Jussi Sopanen

    Research output: Contribution to journalArticleScientificpeer-review

    Abstract

    Sensor placement is a vital factor affecting the quality and accuracy of virtual sensing. Modal expansion techniques are well-known methods to expand the measured displacements or accelerations to all unmeasured degrees of freedom. For this purpose, a two-phase sensor placement optimisation method is proposed for commonly used triaxial accelerometers. The method uses minimum variance criterion of an estimation error of structural responses. A measure of redundancy of information is introduced as an additional criterion for the placement of the triaxial sensors to minimise the redundancy between the sensors. This was addressed to avoid spatial correlation and clustering of the sensor locations. In addition, a proposal for modal displacement-based weighting is introduced to avoid potential selection of sensor locations with low vibration energy, which can be critical in noisy environments. The efficiency of the proposed method is verified with numerical models of different types of structures and finally with the laboratory scale experiments. The mean error of the reconstructed response in this particular experimental case study was 1.4% of the maximum measured response amplitude. This method is especially applicable to large finite element models of industrial-scale structures with fine meshes.

    Original languageEnglish
    Article number109581
    Number of pages22
    JournalMechanical Systems and Signal Processing
    Volume184
    DOIs
    Publication statusE-pub ahead of print - 1 Feb 2023
    MoE publication typeA1 Journal article-refereed

    Keywords

    • Information redundancy
    • Optimal sensor placement
    • Structural health monitoring

    Fingerprint

    Dive into the research topics of 'Optimal sensor placement of triaxial accelerometers for modal expansion'. Together they form a unique fingerprint.

    Cite this