Acta Photonica Sinica, Volume. 51, Issue 4, 0406001(2022)
Feature Extraction Algorithm of Optical Fiber Vibration Signal Based on Compensation Distance Estimation
Figures & Tables(10)
Fig. 4. Classification accuracy under different dimensionality reduction dimensions of three kinds of signals
Fig. 7. Accuracy and accuracy difference under different signal-to-noise ratio
Table 1. Classification accuracy of three types of signals
View tableView in ArticleTable 1. Classification accuracy of three types of signals
Disturbance signal Number of tests Correct number Number of errors Accuracy Shaking 20 20 0 100% Tapping 20 15 5 75% Knocking 20 18 2 90%
Table 2. Characteristic dimensions at different thresholds
View tableView in ArticleTable 2. Characteristic dimensions at different thresholds
Coefficient N 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 0.55 0.6 0.7 Dimension 96 73 61 53 46 37 37 30 24 18 15 10
Table 3. Classification accuracy under different signal-to-noise ratio
View tableView in ArticleTable 3. Classification accuracy under different signal-to-noise ratio
SNR/ dB MFCC MFCC+PCA Proposed algorithm 18 24 30 18 24 30 20 78% 87% 85% 82% 88% 87% 90% 15 73% 87% 87% 85% 80% 83% 85% 10 72% 85% 83% 80% 80% 83% 87% 5 65%
63%
80% 78% 75% 78% 83% 87% 0 78% 73% 72% 77% 78% 82%
Tools
Get Citation
Copy Citation Text
Haiyan XU, Qingkang KOU, Yingjuan XIE, Jun ZHU, Min LI. Feature Extraction Algorithm of Optical Fiber Vibration Signal Based on Compensation Distance Estimation[J]. Acta Photonica Sinica, 2022, 51(4): 0406001
Paper Information
Category: Fiber Optics and Optical Communications
Received: Sep. 14, 2021
Accepted: Oct. 26, 2021
Published Online: May. 18, 2022
The Author Email: XU Haiyan (hyxu@hhu.edu.cn)
© Copyright 2018-2021 | Chinese Laser Press.
All Rights Reserved 沪ICP备15018463号-20