Modal parameter identification using the continuous wavelet transform

Γιάννη, Ήρα; Gianni, Ira

dc.contributor.author	Γιάννη, Ήρα	el
dc.contributor.author	Gianni, Ira	en
dc.date.accessioned	2022-07-13T08:56:12Z
dc.date.available	2022-07-13T08:56:12Z
dc.identifier.uri	https://dspace.lib.ntua.gr/xmlui/handle/123456789/55383
dc.identifier.uri	http://dx.doi.org/10.26240/heal.ntua.23081
dc.description	Εθνικό Μετσόβιο Πολυτεχνείο--Μεταπτυχιακή Εργασία. Διεπιστημονικό-Διατμηματικό Πρόγραμμα Μεταπτυχιακών Σπουδών (Δ.Π.Μ.Σ.) “Δομοστατικός Σχεδιασμός και Ανάλυση των Κατασκευών”	el
dc.rights	Default License
dc.subject	Ιδιοχαρακτηριστικά	el
dc.subject	Επεξεργασία σημάτων	el
dc.subject	Συνεχής κυματιδιακός μετασχηματισμός	el
dc.subject	Κυματιδιακή ανάλυση	el
dc.subject	Κυματίδια	el
dc.subject	Signal processing	en
dc.subject	Wavelet transform	en
dc.subject	Continuous wavelet transform	en
dc.subject	Modal parameter	en
dc.subject	Wavelet	en
dc.subject	Morlet wavelet	el
dc.subject	Cauchy wavelet	el
dc.subject	Ridge	el
dc.title	Modal parameter identification using the continuous wavelet transform	en
heal.type	masterThesis
heal.classification	Signal Processing	en
heal.language	en
heal.access	free
heal.recordProvider	ntua	el
heal.publicationDate	2022-03-02
heal.abstract	Modal parameter identification is a fundamental procedure in structural engineering, applied both in the design phase of new structures as well as the monitoring or the assessment of existing structures. The objective of this thesis is the application of the Continuous Wavelet Transform (CWT) on the free decay responses of damped linear structural systems, with the aim of the identification of their modal parameters. Both single degree of freedom (s.d.o.f.) and multi degree of freedom (m.d.o.f.) systems were examined, under the assumptions of weak damping and of linear behaviour. The condition of weak damping allowed the consideration of the processed signals as a sum of components consisting of asymptotic frequency modulated signals. Consequently, the concept of the analytic signal with the terms of instantaneous amplitude, phase and frequency was applied and the modal parameter identification was performed directly, by employing procedures and equations involving these signals. The CWT is a multi-scale time-frequency signal processing method, based on a set of family wavelets formed by scaling and translation of a prototype mother wavelet. Two complex mother wavelets were employed, the Complex Morlet wavelet and the Cauchy wavelet of order n. The CWT using a complex mother wavelet returns information about both the instantaneous amplitude and instantaneous phase of each component within the processed signals, as the calculated data tends to “concentrate” near a series of curves in the time-frequency domain, called the ridges of the transform. Ridges can be defined as the place where the instantaneous frequency of the signal is equal to the analyzing wavelet’s center frequency. Each ridge corresponds to a component of the signal, and thus, the identification of the ridges allows for the estimation of the corresponding instantaneous frequencies, the damping rations and the mode shapes. The algorithms implemented for the identification of ridges focus on differential methods, specifically the “Simple” method that employs the modulus of the CWT and the “Marseille” method that employs the phase of the CWT. The thesis is structured into three thematic sections: The first thematic section presents the essential engineering, physical and mathematical theoretical background, with emphasis in signal processing techniques and the fundamental definitions and properties of the CWT. The second thematic section focuses on the methods and algorithms that are applied for modal parameter identification using the CWT and the third involves two numerical applications of the provided methods, one over a s.d.o.f system and one on a m.d.o.f. system. The implementation of the algorithms, the calculations and the visualizations are done with MATLAB ver.R2021a, using the standard features and additionally the Signal Processing Toolbox and the Wavelet Toolbox.	en
heal.advisorName	Φραγκιαδάκης, Μιχαήλ	el
heal.committeeMemberName	Φραγκιαδάκης, Μιχαήλ	el
heal.committeeMemberName	Παπαδόπουλος, Βησσαρίων	el
heal.committeeMemberName	Τριανταφύλλου, Σάββας	el
heal.academicPublisher	Εθνικό Μετσόβιο Πολυτεχνείο. Σχολή Πολιτικών Μηχανικών	el
heal.academicPublisherID	ntua
heal.numberOfPages	163 σ.	el
heal.fullTextAvailability	false
heal.fullTextAvailability	false