Iterative pole-zero finite element model updating using generic parameters. (November 2018)

• Record Type: Journal Article
• Authors: Dorosti, M.; Fey, R.H.B.; Heertjes, M.F.; van de Wal, M.M.J.; Nijmeijer, H.
• Is Part Of: Mechatronics. Volume 55(2018)
• Page Start: 180
• Page End: 193
• Publication Date: 2018-11
• Subjects: FE model -- Model updating -- System identification -- Prediction and control -- Generic parameters; Computer integrated manufacturing systems -- Periodicals; Flexible manufacturing systems -- Periodicals; Mechatronics -- Periodicals; Productique -- Périodiques; Fabrication, Systèmes flexibles de -- Périodiques; Mécatronique -- Périodiques; Computer integrated manufacturing systems; Flexible manufacturing systems; Mechatronics; Periodicals; 629.89
• DOI: 10.1016/j.mechatronics.2018.06.012
• Languages: English
• ISSNs: 0957-4158

Abstract

Abstract: Prediction and control of unmeasurable performance variables in complex large-scale systems is an important topic in systems and control. The preferred solution to this problem is to have a relatively low-order and accurate standard plant model. A reduced Finite Element (FE) dynamic model may be a good candidate for that purpose apart from the fact that its accuracy is always limited, since it never truly matches the real structure. In this paper, an iterative procedure is proposed to update the FE model using poles and zeros estimated from the measured Frequency Response Functions (FRFs). Assuming the errors in the FE stiffness and/or damping matrix to be dominant over the mass matrix, but the exact physical parameters responsible for these errors to be unknown, the eigenvalues of the (sub)structure stiffness and/or damping matrix may be updated to reduce the differences between the poles and zeros from the model and measured FRFs. The proposed procedure matches poles and zeros sequentially in an iterative manner as to realize convergence to the measured quantities. This procedure, further referred to as Iterative Pole-Zero (IPZ) model updating, facilitates the process of obtaining improved parametric models which can be used for accurate prediction and control of unmeasurable performance variables. The proposed procedure is first illustrated using a mass-spring model and then is validated via a collocated FRF measurement from an experimental beam setup.