Determination of wavelet ridges of nonstationary signals by singular value decomposition

Creative Commons License

Ozkurt N., Savaci F.

IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS II-EXPRESS BRIEFS, vol.52, no.8, pp.480-485, 2005 (SCI-Expanded) identifier identifier

  • Publication Type: Article / Article
  • Volume: 52 Issue: 8
  • Publication Date: 2005
  • Doi Number: 10.1109/tcsii.2005.849041
  • Journal Name: IEEE TRANSACTIONS ON CIRCUITS AND SYSTEMS II-EXPRESS BRIEFS
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus
  • Page Numbers: pp.480-485
  • Keywords: instantaneous frequency, singular value decomposition (SVD), wavelet ridge, FREQUENCY, RECONSTRUCTION, TRANSFORM, DYNAMICS
  • Dokuz Eylül University Affiliated: No

Abstract

The ridges obtained from chaotic signals can give the relevant information about the phase structures of the dynamical systems. Therefore, a new wavelet ridge determination method for the noisy signals and nonstationary signals, which is based on the singular value decomposition (SVD) has been proposed in this paper. The proposed method has been compared with Carmona method for monocomponent signals, and multicomponent signals. The proposed method is computationally more effective than the Carmona method to determine the actual ridges. Also, the ridges of the periodic limit cycles and chaotic attractors have been determined by using the SVD-based method to find the degree of chaoticity.