An Adaptive Graph Spectral Analysis Method for Feature Extraction of an EEG Signal

2019 ◽  
Vol 19 (5) ◽  
pp. 1884-1896 ◽  
Author(s):  
Shanzhi Xu ◽  
Hai Hu ◽  
Linhong Ji ◽  
Peng Wang
Keyword(s):  
Feature Extraction ◽  
Spectral Analysis ◽  
Eeg Signal ◽  
Analysis Method ◽  
Spectral Analysis Method

scholarly journals Residual Network Structure-based High Accuracy Spectral Analysis Method

2019 ◽  
Vol 15 (11) ◽  
pp. 3090
Author(s):  
Wu Sai ◽  
Wang Zhihui ◽  
Meng Sachura ◽  
Zheng Weijun ◽  
Shao Weiping
Keyword(s):  
Spectral Analysis ◽  
Network Structure ◽  
High Accuracy ◽  
Analysis Method ◽  
Residual Network ◽  
Spectral Analysis Method

scholarly journals Spectral whitening based on the singular spectral analysis method

2018 ◽  
Vol 15 (4) ◽  
pp. 1460-1469 ◽  
Author(s):  
Rafael R Manenti ◽  
Wilker E Souza ◽  
Milton J Porsani
Keyword(s):  
Spectral Analysis ◽  
Analysis Method ◽  
Singular Spectral Analysis ◽  
Spectral Analysis Method

Comparison of Hydraulic Pump Faults Diagnosis Methods: Wavelet vs. Spectral Analyses

2005 ◽  
Author(s):  
Yingjie Gao ◽  
Qin Zhang ◽  
Xiangdong Kong
Keyword(s):  
Spectral Analysis ◽  
Wavelet Transform ◽  
Finite Length ◽  
Data Sets ◽  
Analysis Method ◽  
Hydraulic Pump ◽  
Transform Method ◽  
Outlet Pressure ◽  
Spectral Analysis Method ◽  
Faults Diagnosis

This paper introduces two faults diagnosis methods, a conventional spectral analysis method and a wavelet transform method, for hydraulic pump applications. The fundamental technologies of both methods, as well as their performance in detecting a few common hydraulic pump defects, are described in this paper. The performance of both diagnoses methods were evaluated based on experimental results. In order to eliminate the effects of border distortion arising from applying wavelet transform to finite-length signals, the pump outlet pressure in this case, a preprocess on the obtained signals is carried to clean up the errors prior to faults diagnosis analysis. Validation results obtained from both methods in analyzing the same data sets indicated that the wavelet transform based method showed a more sensitive and robust detecting capability than that obtained from a spectrum analyses approach.

Sign in / Sign up

Export Citation Format

Share Document