Premium
Pattern recognition based on enhanced multifeature parameters for vibration events in φ‐OTDR distributed optical fiber sensing system
Author(s) -
Xu Chengjin,
Guan Junjun,
Bao Ming,
Lu Jiangang,
Ye Wei
Publication year - 2017
Publication title -
microwave and optical technology letters
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.304
H-Index - 76
eISSN - 1098-2760
pISSN - 0895-2477
DOI - 10.1002/mop.30886
Subject(s) - optical time domain reflectometer , vibration , signal (programming language) , time domain , acoustics , optical fiber , noise (video) , signal to noise ratio (imaging) , distributed acoustic sensing , fiber , spectral density , computer science , fiber optic sensor , electronic engineering , engineering , artificial intelligence , materials science , telecommunications , physics , fiber optic splitter , computer vision , composite material , image (mathematics) , programming language
A new multifeature recognition method for optical fiber vibration signals is proposed to solve the difficulty with the present long‐distance distributed optical fiber sensing systems based on phase‐sensitive optical time‐domain reflectometer (φ‐OTDR) in effectively distinguishing vibrational intrusions. Firstly, wide‐band background noise is reduced by spectral subtraction to enhance time‐frequency characteristics of vibration signals. Then, multifeature parameters including short‐time energy ratio, short‐time level crossing rate, vibration duration, and power‐spectrum energy ratio are extracted from the vibration signals. Finally, by utilizing support vector machine classifiers to identify multifeature vectors of different types of vibration signals, patterns of vibration events are recognized accurately. Experiments show that after using this method to process 800 vibration signal samples generated by four different vibration events, namely, taping, striking, shaking, and crushing, the signal‐to‐noise ratio is enhanced by more than 10 dB and the recognition rates of vibration events are over 90%. The experimental results prove that this method can improve the classification accuracy of vibration events detected by φ‐OTDR distributed optical fiber sensing systems with a recognition time below 0.6 s, especially for long distances.