Share Email Print
cover

Proceedings Paper

A novel fiber-optical vibration defending system with on-line intelligent identification function
Author(s): Huijuan Wu; Xin Xie; Hanyu Li; Xiaoyu Li; Yu Wu; Yuan Gong; Yunjiang Rao
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

Capacity of the sensor network is always a bottleneck problem for the novel FBG-based quasi-distributed fiberoptical defending system. In this paper, a highly sensitive sensing network with FBG vibration sensors is presented to relieve stress of the capacity and the system cost. However, higher sensitivity may cause higher Nuisance Alarm Rates (NARs) in practical uses. It is necessary to further classify the intrusion pattern or threat level and determine the validity of an unexpected event. Then an intelligent identification method is proposed by extracting the statistical features of the vibration signals in the time domain, and inputting them into a 3-layer Back-Propagation(BP) Artificial Neural Network to classify the events of interest. Experiments of both simulation and field tests are carried out to validate its effectiveness. The results show the recognition rate can be achieved up to 100% for the simulation signals and as high as 96.03% in the real tests.

Paper Details

Date Published: 15 October 2013
PDF: 5 pages
Proc. SPIE 8924, Fourth Asia Pacific Optical Sensors Conference, 89241P (15 October 2013); doi: 10.1117/12.2034275
Show Author Affiliations
Huijuan Wu, Univ. of Electronic Science and Technology of China (China)
Xin Xie, Univ. of Electronic Science and Technology of China (China)
Hanyu Li, Univ. of Electronic Science and Technology of China (China)
Xiaoyu Li, Univ. of Electronic Science and Technology of China (China)
Yu Wu, Univ. of Electronic Science and Technology of China (China)
Yuan Gong, Univ. of Electronic Science and Technology of China (China)
Yunjiang Rao, Univ. of Electronic Science and Technology of China (China)


Published in SPIE Proceedings Vol. 8924:
Fourth Asia Pacific Optical Sensors Conference
Minghong Yang; Dongning Wang; Yun-Jiang Rao, Editor(s)

© SPIE. Terms of Use
Back to Top