Real-time Compression Method Of Massive Distributed Fiber Sensing Signals

Distributed optical fiber sensing technology can continuously measure the external physical parameters along the optical fiber and pick up the temporal and spatial information of the physical parameters.Due to the characteristics of corrosion resistance,anti-electromagnetic interference and remote telemetry,distributed optical fiber sensing technology occupies a very important position in the field of infrastructure safety monitoring,including the application in the energy industry,electrical power industry,aviation,architecture,communication,transportation and security field.However,tens of thousands of data will be generated by distributed optical fiber sensing system in large-scale,long-distance and continuous monitoring.There will be a great pressure for the massive data storage,transmission and processing.Data compression is commonly used to solve the problem of massive data storage and transmission as an effective way.In this thesis,the following studies have been conducted around the data compression method of distributed optical fiber sensing signals.1.Based on the statistical method of the box plot,an effective signal detection method for distributed fiber-optic vibration sensor signals is proposed,which can reduce the 97.2 percent of invalid data of background environment.2.Based on Otsu threshold method and LWT,a data real-time compression method is proposed.LWT has higher calculation efficiency than the conventional DWT and DCT.In allusion to three typical Distributed optical fiber sensing signals,including POTDR signal of length 7200,?-OTDR signal of length 50001 and BOTDA signal of length 40,the performances of LWT compression method in compression time are10^-3s,10^-2s,and 10^-4s,the performances of DWT compression method in compression time are 10^-1s,10¹s,and 10^-3s,the performances of DCT compression method in compression time are 10⁰s,10²s,and 10^-4s.Moreover,compared with conventional hard and soft threshold method,the Otsu threshold method has good adaptability to the environment.3.The proposed data compression method can improve the storage efficiency of distributed fiber sensing signals on big data processing platform.After LWT real-time compression,the storage time on big data processing platform is reduced to about one fifteenth of the storage time without compression.4.Based on the LWT's compression method on other distributed fiber sensing signal experiments,the feasibility of LWT's compression method extending to other distributed fiber sensing fields is demonstrated.LWT-based compression method shows its advantages in real time performance and signal fidelity for POTDR signal.LWT-based compression method shows its advantages in real time performance,compression ratio and signal fidelity for?-OTDR signal.LWT-based compression method shows its advantages in BFS error for BOTDA signal.