Identification of Nonlinear Systems
At first glance few problems arise in the area of structural dynamics which cannot be solved by means of today’s experimental and analytical tools. Thus, the elastodynamic characteristics can be determined by using common experimental or analytical method
- PDF / 23,897,523 Bytes
- 512 Pages / 481.89 x 691.654 pts Page_size
- 67 Downloads / 209 Views
IDENTIFICATION OF VIBRATING STRUCTURES
EDITED BY
H.G. NATKE UNIVERSITY OF HANNOVER
SPRINGER-VERLAG WIEN GMBH
This work is subject to copyright. All rights are reserved, whether the whole or part of the material is concerned specifically those of translation, reprinting, re-use of illustrations, broadcasting, reproduction by photocopying machine or similar means, and storage in data banks.
©
1982 by Springer-Verlag Wien
Originally published by Springer-Verlag Wien New York in 1982
ISBN 978-3-211-81651-6 DOI 10.1007/978-3-7091-2896-1
ISBN 978-3-7091-2896-1 (eBook)
PREFACE
The dynamic behaviour of novel and complicated structures often needs to be investigated by system analysis and system identification, since it usually has to meet certain requirements. A priori knowledge concerning the real system is gained by system analysis and/or previous tests, and it results in a non-parametric and/or in a parametric mathematical model. The identification of system parameters, i.e. experimental system analysis, is performed using measured quantities and taking into account deterministic and stochastic errors. If results of the identification have to be compared with the results of the system analysis, and if certain error bounds are exceeded, the model has to be improved. System identification has to take into account random aspects (errors, test signals), the real dynamic behaviour (damping coupling, non-linearities) and questions concerning practical handling (including large systems, economics). A broad understanding of system identification needs as its basis an extended theory of structural vibrations and estimation (stochastic processes), and must be coupled with practical aspects including experience and validated software. The course on Identification of Vibrating Structures, the lecture notes of which are collected in this volume, deals with the topics mentioned above. First an introduction into the subject is given, and the theoretical background of vibrating structures and parameter estimation methods is dealt with. The following lectures deal with several identification methods including applications. The next papers discuss the indirect identification, that means the adjustment of theoretical models (results of system analysis) by the results of vibration tests (estimated values). These parts are supplemented by a presentation of an example of commercially available hard- and software including applications. For the
practical application of system identification concerning large and complicated structures synthesizing techniques (substructure techniques) including error analysis are necessary, which are the subject of the following two papers, followed by identification of non-linear systems. The closing lectures deal with modern developments, firstly from the point of view of control theory, and secondly by coming from the theory of stochastic systems. The last three lectures deal with theoretical aspects, including examples of (simple) systems mainly with regard to non-deterministic system analysis
