A method to eliminate riding waves appearing in the empirical AM/FM demodulation

Authors:
Zhijing Yang;Lihua Yang;Chunmei Qing;Daren Huang
Affiliations:
School of Mathematics and Computing Science, Sun Yat-sen University, Guangzhou 510275, PR China;School of Mathematics and Computing Science, Sun Yat-sen University, Guangzhou 510275, PR China;School of Informatics, University of Bradford, Bradford BD7 1DP, UK;School of Mathematics and Computing Science, Sun Yat-sen University, Guangzhou 510275, PR China
Venue:
Digital Signal Processing
Year:
2008

Citing 5
Cited 2

Time-frequency analysis: theory and applications

Time-frequency analysis: theory and applications
Signal Period Analysis Based on Hilbert-Huang Transform and Its Application to Texture Analysis

ICIG '04 Proceedings of the Third International Conference on Image and Graphics
An EMD-based recognition method for Chinese fonts and styles

Pattern Recognition Letters
A novel pitch period detection algorithm based on hilbert-huang transform

SINOBIOMETRICS'04 Proceedings of the 5th Chinese conference on Advances in Biometric Person Authentication
On the analytic signal, the Teager-Kaiser energy algorithm, andother methods for defining amplitude and frequency

IEEE Transactions on Signal Processing

An empirical re-sampling method on intrinsic mode function to deal with speed variation in machine fault diagnostics

Applied Soft Computing
An improved envelope algorithm for eliminating undershoots

Digital Signal Processing

Quantified Score

Hi-index	0.00

Visualization

Abstract

Empirical mode decomposition can decompose a complicated signal into a sum of finite intrinsic mode functions whose instantaneous frequencies computed by the analytic signal method give a physically meaningful characterization of the signal. However, it is found that computing the instantaneous frequency of an intrinsic mode function through the Hilbert transform may produce some ridiculous results. Recently, an empirical AM/FM (amplitude and frequency modulation) demodulation for computing the instantaneous frequency of an intrinsic mode function has been proposed by Huang and his coworkers. This method totally eschews the Hilbert transform and can compute the instantaneous frequency of an intrinsic mode function very accurately. However, this paper will show that the empirical FM part extracted through the empirical AM/FM demodulation may contain riding waves and is no longer an intrinsic mode function. That will make the instantaneous frequency nonsensical. To overcome this drawback, an improved method, called riding wave turnover-empirical AM/FM demodulation, is proposed in this paper. Experiments show very positive results.