Multi-Point Detection For Polarization-Sensitive Optical Time Domain Reflectometry And Its Applications In Electric Power Industry
Huijuan Wu,Jingwu Luo,Jiang Wu,Jun Liu,Lidong Lv,Yunjiang Rao,Xiaoyan Sun,David Atubga
DOI: https://doi.org/10.1109/PIERS.2016.7735791
2016-01-01
Abstract:In this paper, we propose a novel method of disturbance detection by using image morphology to solve the challenging multi-point detection problems in polarization-sensitive optical time domain reflectometry (P-OTDR), which restrict its practical applications greatly. The new methodology is processed by accumulating the temporally differentiated OTDR traces and building up a two-dimensional temporal-spatial SOP evolving graph, and then using edge detection and automatic-clustering, adopted from imaging processing and statistical classification techniques, to distinguish different disturbance points and pinpoint their locations. The proposed method proved better performance than the conventional direct differentiation method and the Fast Fourier Transform (FFT) spectrum analysis. The location accuracy in particular, has improved significantly. Also, we present two typical field applications of this representative Distributed Optical Fiber Sensor (DOFS) in the electric power industry, for on-line safety monitoring of both the long-distance power transmission cables buried underground and those overhead. It demonstrated that P-OTDR is a promising technology for on-line safety monitoring of long-distance power transmission lines and optical fiber communication cables, due to its outstanding advantages, such as fully passive operation, electro-magnetic interference immunity (EMI), corrosion resistance, good reliability, and more especially its long-distance single-end detection ability, quantitative measurement capability and moderate sensibility for disturbance detection.