Font Size: a A A

Identification And Detection Of Vibration Signal In Optical ?-OTDR System

Posted on:2021-06-15Degree:MasterType:Thesis
Country:ChinaCandidate:C D LiuFull Text:PDF
GTID:2480306554466584Subject:Master of Engineering
Abstract/Summary:PDF Full Text Request
Single-mode fiber(SMF)has better sensing property in phase-sensitive optical time-domain reflectometer(?-OTDR),but the deficiencies are listed as follows:(1)the capture fraction of reflected light and the nonlinear threshold level of SMF is low,which causes the low power of pulsed lights that can be injected and weak signal of collected Rayleigh scattered light,limiting the sensing distance;(2)only one mode of optical signal can be transmitted in SMF,which cannot make up for the sudden changes of vibration peak caused by noises due to the great influence of the inhomogeneity of fiber medium and the mismatching of polarization state.Few-mode fiber(FMF)is the transitional fiber of SMF and multi-mode fiber(MMF),of which fiber core radius is between them.FMF has a higher nonlinear threshold level compared with SMF in the field of optical fiber sensing,which can be injected to the exploration pulse with the higher energy to obtain the stronger Rayleigh backscatter light for the promotion of the sensing distance.A variety of modes of light can be transmitted by few-mode fiber(FMF),the influence of single mode light wave caused by noise can be reduced by means of the combination for the signals of different modes.The vibration sensing property of two-mode fiber(TMF)are analyzed,and the new algorithm of the vibration signal detection and recognition in allusion to?-OTDR system is proposed,the specific details are listed as follows:(1)The research on the mode theory of optical fiber is conducted and the power distribution of Rayleigh backscatter signals in each mode of the FMF is deduced.The research on the?-OTDR is conducted,and the advantages of coherent detection technology and orthogonal demodulation in the system are deduced.At last,analyze the main parameters that affect the system performance of?-OTDR system.(2)The different modes of exploration lights are injected to the dual-mode TMF by the photon lantern,at the same time,the Rayleigh backscatter lights with the same light mode are collected,the average signal-to-noise ratio(SNR)of 2.99d B and 2.85d B is obtained by the optical signal with the modes of LP01 and LP11a.The method of setting the signal compensation with different sampling rate for different modes is proposed in allusion to the different velocities of propagation of different modes of lights in the optical fiber,the signals of the two modes after the signal compensation are integrated and the SNR is increased by about 7.56d B.(3)The TMF is adopted as the optical fiber with large effective area,?-OTDR system is connected in the quasi-single mode and the number of vibration sources on the optical fiber is increased,the sensing performance of TMF to multi-vibration under the quasi-single mode operation is verified by means of the experiment.(4)The method of amplitude difference accumulation is adopted by?-OTDR system to determine the position of vibration signal,there are the problems such as the limited frequency range to be detected and the modification of N value for vibration signals at different frequencies in the system.The interference curve of the Rayleigh backscatter lights is processed by means of the method of calculating dispersion in statistics to solve the problems,which promotes the accuracy of vibration peak detection and the range of detectable frequency.(5)The research on the identification of vibration signal is conducted based on the combination of empirical mode decomposition(EMD)and neural network.EMD is adopted for the decomposition and feature extraction of the time-domain signals of vibration points.And the features,variance and range of the signals are input into the neural network for training,the recognition accuracy for 40Hz vibration signals can be reached to 96.49%in the end.
Keywords/Search Tags:two-mode fiber, phase-sensitive optical time-domain reflectometer, empirical mode decomposition, neural network
PDF Full Text Request
Related items