Wavelet-based detection of abrupt changes in natural frequencies of time-variant systems. (December 2015)

• Record Type: Journal Article
• Authors: Dziedziech, K.; Staszewski, W.J.; Basu, B.; Uhl, T.
• Is Part Of: Mechanical systems and signal processing. Volume 64/65(2015)
• Page Start: 347
• Page End: 359
• Publication Date: 2015-12
• Subjects: System identification -- Time-variant systems -- Natural frequency -- Abrupt change detection -- Wavelets; Structural dynamics -- Periodicals; Vibration -- Periodicals; Constructions -- Dynamique -- Périodiques; Vibration -- Périodiques; Structural dynamics; Vibration; Periodicals; 621
• DOI: 10.1016/j.ymssp.2015.03.012
• Languages: English
• ISSNs: 0888-3270

Abstract

Abstract: Detection of abrupt changes in natural frequencies from vibration responses of time-variant systems is a challenging task due to the complex nature of physics involved. It is clear that the problem needs to be analysed in the combined time–frequency domain. The paper proposes an application of the input–output wavelet-based Frequency Response Function for this analysis. The major focus and challenge relate to ridge extraction of the above time–frequency characteristics. It is well known that classical ridge extraction procedures lead to ridges that are smooth. However, this property is not desired when abrupt changes in the dynamics are considered. The methods presented in the paper are illustrated using simulated and experimental multi-degree-of-freedom systems. The results are compared with the classical Frequency Response Function and with the output only analysis based on the wavelet auto-power response spectrum. The results show that the proposed method captures correctly the dynamics of the analysed time-variant systems. Highlights: A new method for detection of abrupt changes in natural frequencies is proposed. The method uses the input–output algorithm that utilizes the wavelet-based Frequency Response Function (FRF). The major focus is on ridge extraction and occupation measures applied to the FRF characteristics and the estimation of time instants for the abrupt changes in natural frequencies. The proposed methodology is illustrated using simulated multi-degree-of-freedom data and experimental input–output data from of a three-floor building structure. The method captures the dynamics of the analysed simulated and experimental time-variant vibration systems. All abrupt changes of natural frequency are identified correctly.