Non-Parametric Identification of Asymmetric Signals and Characterization of a Class of Non-Linear Systems Based on Frequency Modulation

Václav Ondra, Ibrahim A. Sever, Christoph W. Schwingshackl

Research output: Contribution to conferenceConference Paperpeer-review

Abstract

Non-parametric and parametric identification of a non-linear system is often performed by estimating instantaneous amplitude and frequency using the Hilbert transform. However, the Hilbert transform cannot be used for the accurate analysis of asymmetric signals and the reliable estimation of intra-wave frequency modulation. This paper proposes two alternatives to the Hilbert transform which not only avoid some of its mathematical and numerical issues, but also allow the above mentioned analyses. The first method, based on zero-crossing, allows the backbone and damping curves as well as the elastic and damping force characteristics of an asymmetric free decay to be identified. The application and accuracy of this method are demonstrated on the free decay of the system with off-centre clearance. The second method, based on direct quadrature, estimates intrawave frequency modulation frequency with sufficient resolution for characterization of non-linear systems which have stiffness non-linearities. The use of this method is shown on a system with cubic hardening stiffness.
Original languageEnglish
Number of pages10
DOIs
Publication statusPublished - 2016
EventASME International Mechanical Engineering Congress and Exposition, IMECE 2016 - Phoenix, United States
Duration: 10 Nov 2016 → …

Conference

ConferenceASME International Mechanical Engineering Congress and Exposition, IMECE 2016
Abbreviated titleIMECE 2016
Country/TerritoryUnited States
CityPhoenix
Period10/11/16 → …

Fingerprint

Dive into the research topics of 'Non-Parametric Identification of Asymmetric Signals and Characterization of a Class of Non-Linear Systems Based on Frequency Modulation'. Together they form a unique fingerprint.

Cite this