[104] Invited Overview Presentation
A fourth-order synchroextracting transform for the analysis of non-stationary signals with fast varying instantaneous frequency Kun Yu1,2 and Tian Ran Lin2
1School of Mechanical Engineering and Automation, Northeastern University, Shenyang, Liaoning 110819,
PR China
2School of Mechanical and Automotive Engineering, Qingdao University of Technology, Qingdao, Shandong 266520, PR China.
Email: trlin@qut.edu.cn
Time-frequency analysis can provide a better understanding of the complex structure in a non-stationary signal. As a post-processing algorithm of the short-time Fourier transform (STFT), the synchroextracting transform can substantially improve the readability of the time-frequency representation (TFR) and signal reconstruction of a non-stationary signal. A synchroextracting transform works quite well when dealing with weak frequency-modulated signals, though the time-frequency representation of a signal becomes blurry when the signal has a strong frequency modulation characteristic. To address this issue, a new instantaneous frequency estimation technique is proposed in this study to accurately capture the rapid variation of instantaneous frequency in a signal. In this approach, both the amplitude and phase of a non-stationary signal are expanded into a fourth-order Taylor expansion and then the higher order meaningless terms are subtracted from the signal to obtain the desired instantaneous frequency estimator. A fourth-order synchroextracting transform algorithm is then proposed by binding together the great estimation capacity of the fourth-order instantaneous frequency estimation method and the excellent energy extraction capability of a synchroextracting operator. It is shown in the study that the fourth-order synchroextracting transform algorithm can improve the readability of the time-frequency representation of a strong frequency-varying non-stationary signal and enable an accurate signal reconstruction.
Keywords: bearing fault diagnosis, non-stationary signal, instantaneous frequency estimation method, synchroextracting operator, fourth-order synchroextracting transform.
2School of Mechanical and Automotive Engineering, Qingdao University of Technology, Qingdao, Shandong 266520, PR China.
Email: trlin@qut.edu.cn
Time-frequency analysis can provide a better understanding of the complex structure in a non-stationary signal. As a post-processing algorithm of the short-time Fourier transform (STFT), the synchroextracting transform can substantially improve the readability of the time-frequency representation (TFR) and signal reconstruction of a non-stationary signal. A synchroextracting transform works quite well when dealing with weak frequency-modulated signals, though the time-frequency representation of a signal becomes blurry when the signal has a strong frequency modulation characteristic. To address this issue, a new instantaneous frequency estimation technique is proposed in this study to accurately capture the rapid variation of instantaneous frequency in a signal. In this approach, both the amplitude and phase of a non-stationary signal are expanded into a fourth-order Taylor expansion and then the higher order meaningless terms are subtracted from the signal to obtain the desired instantaneous frequency estimator. A fourth-order synchroextracting transform algorithm is then proposed by binding together the great estimation capacity of the fourth-order instantaneous frequency estimation method and the excellent energy extraction capability of a synchroextracting operator. It is shown in the study that the fourth-order synchroextracting transform algorithm can improve the readability of the time-frequency representation of a strong frequency-varying non-stationary signal and enable an accurate signal reconstruction.
Keywords: bearing fault diagnosis, non-stationary signal, instantaneous frequency estimation method, synchroextracting operator, fourth-order synchroextracting transform.
