| Distributed optical fiber sensing is a sensing technology which utilizes optical fiber as sensing medium to realize real-time monitoring of disturbance along the sensing fiber.Due to its advantages such as immunity to electromagnetic interference,high detection sensitivity,geometric versatility and low cost,it is especially suitable for long-distance and large-scale monitoring tasks.At present,it has been widely used in the fields including the health monitoring of civil buildings,the safety monitoring of oil pipelines,national border security.Distributed optical fiber sensing system has many classifications according to the different structure and sensing principle.The ?-OTDR system has become one of the hotspots in the field of distributed optical fiber sensing because of its excellent distributed performance and high frequency response range.In this thesis,the research subject is the distributed optical fiber vibration sensing system based on the?-OTDR and the aim of our works is to reduce the background noise in the system.Through investigating and analyzing the theory of ?-OTDR and correlative signal processing algorithm,a series of signal processing methods which can improve the signal-to-noise ratio of ?-OTDR system are put forward.The main research contents and results are as follows:1?We introduce the background and significance of distributed optical fiber sensing technology briefly.Meanwhile the advantages and disadvantages of distributed optical fiber sensing systems based on different configuration are presented.Furthermore,the development and status of ?-OTDR system are described in detail.We discuss the working principle of OTDR and(p-OTDR system which lays a foundation for the follow-up study.2?A signal processing method of ?-OTDR system based on curvelet transform algorithm is proposed to reduce the time domain noise for enhancing the detection performance.The Gray image consisted of Rayleigh backscatter traces is decomposed into the curvelet coefficients of different scales and directions by curvelet transform,and the noise reduction is processed by the Monte Carlo threshold criterion.In view of the actual denoising effect,the threshold scheme is optimized.Experiment results show that the SNR of location information is increased to 15.6dB and 16dB respectively for 50Hz and 1 kHz vibration events when fiber length is 4km.3?This thesis introduces a novel signal processing method of ?-OTDR system based on the compressive sensing algorithm.Firstly,the three processes of the compressive sensing are expounded in detail,and the sparse,observation and reconstruction schemes which are suitable for the ?-OTDR system are selected.Then,a threshold rule is established to determine the sparsity of the signal that providing references for subsequent signal reconstruction.Finally,the results of the two-part reconstruction are superimposed according to the shrinkage coefficient and the final denoising result is obtained.The experiment results demonstrate that the SNR for 100 Hz and 1.5 kHz external vibration events caused by PZT is enhanced to 29.14 dB and 25.84 dB respectively with fiber length of 3km.4?We propose a signal processing method of ?-OTDR system based on empirical mode decomposition algorithm.Firstly,the signal characteristics of the ?-OTDR system are discussed which indicates that the data composition is different when there is existing vibration or not.Secondly,according to the frequency characteristic of the IMF components,the data composition is determined by calculating the Pearson correlation coefficient.Finally,two methods are adopted to deal with the data in the two cases.Experimental results show that the vibration location of 100Hz and 1.2 kHz can be detected accurately and the SNR of location information has a remarkable improvement to 42.52 dB and 39.58 dB.The main innovations of this thesis are as follows:1?Rayleigh backscatter curves is regarded as a two-dimensional gray image in the signal processing method of ?-OTDR system based on curvelet transform algorithm which is different from the previous signal processing methods.It is fully considered the relationship between Rayleigh backscattering curves and applied the idea of image denoising related processing which provides a new idea for signal processing of(p-OTDR system,and enriches the signal processing method of ?-OTDR system.2?The signal processing method of ?-OTDR system based on the compressive sensing algorithm can effectively reduce the data size.Meanwhile it can realize the synchronization of collection and processing.This solves the problem that large amount of data gathered by ?-OTDR system in long-distance monitoring,and provides an effective scheme for data storage and transmission.It is very important for the development of long-distance distributed optical fiber sensing.3?In the signal processing method of ?-OTDR system based on empirical mode decomposition algorithm,it is the first time in ?-OTDR system to reference the Rayleigh backscattering signal characteristics for signal processing,which is regarded as the most important principle in noise reduction.It can significantly improve the SNR of the system which is beneficial to enhance the performance of ?-OTDR system in practical application. |
PDF Full Text Request