Abstract

A system based on the use of artificial neural networks allowing discrimination of strain and temperature in a conventional Brillouin optical time domain analyzer setup is presented and demonstrated in this paper. This solution allows to perform an automatic discrimination of both parameters without compromising the complexity or cost of the interrogation unit. The classification results, achieved by considering a preprocessing stage with dimensionality reduction via principal component analysis and spatial filtering, improve those obtained in a previous feasibility study.

© 2018 IEEE

Deep neural networks assisted BOTDA for simultaneous temperature and strain measurement with enhanced accuracy

Biwei Wang, Liang Wang, Nan Guo, Zhiyong Zhao, Changyuan Yu, and Chao Lu
Opt. Express 27(3) 2530-2543 (2019)

Signal processing using artificial neural network for BOTDA sensor system

Abul Kalam Azad, Liang Wang, Nan Guo, Hwa-Yaw Tam, and Chao Lu
Opt. Express 24(6) 6769-6782 (2016)

Real-time dynamic strain sensing in optical fibers using artificial neural networks

Sascha Liehr, Lena Ann Jäger, Christos Karapanagiotis, Sven Münzenberger, and Stefan Kowarik
Opt. Express 27(5) 7405-7425 (2019)

Distributed Brillouin frequency shift extraction via a convolutional neural network

Yiqing Chang, Hao Wu, Can Zhao, Li Shen, Songnian Fu, and Ming Tang
Photon. Res. 8(5) 690-697 (2020)

Few-mode fiber multi-parameter sensor with distributed temperature and strain discrimination

An Li, Yifei Wang, Jian Fang, Ming-Jun Li, Byoung Yoon Kim, and William Shieh
Opt. Lett. 40(7) 1488-1491 (2015)