Presentation
29 December 2022 Research on fault identification technology of sound barrier based on distributed optical fiber sensing (Conference Presentation)
Author Affiliations +
Abstract
Due to green building railway requirements, sound barrier, as an effective noise reduction measure, has been widely constructed. However, the sound barrier may fall off if its components are damaged or aging. It is critical to accurately detect the early faults of sound barrier. But existing technical solutions have low recognition rate and lack of real-time performance. To solve these problems, a sound barrier fault identification method based on phase-sensitive optical time-domain reflectometry (ϕ-OTDR) is proposed. We propose a novel method based on optimized multi-domain features for feature extraction and feature screening to describe intrinsic information of the vibration signal. A field experiment was carried out in the Hu-Hang Railway. A total of 405 sets of data were obtained. With the help of quadratic discriminant classifier and 5-fold cross-validation, the average recognition accuracy is 82.3% even under complex field environments.
Conference Presentation
© (2022) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Yuanyuan Yao, Ningmu Zou, Chi Zhang, Qianyi Gao, Yiming Wan, Ruofan Wang, Shisong Zhao, Feng Wang, Yixin Zhang, and Xuping Zhang "Research on fault identification technology of sound barrier based on distributed optical fiber sensing (Conference Presentation)", Proc. SPIE 12314, Optoelectronic Devices and Integration XI, 123140H (29 December 2022); https://doi.org/10.1117/12.2643963
Advertisement
Advertisement
KEYWORDS
Optical fibers

Optical sensing

Feature extraction

Acoustics

Backscatter

Denoising

Fiber optics sensors

Back to Top