Research On Distributed Optical Fiber Vibration Signal Recognition And Location Algorithms

For a long time,the issue of perimeter security has become a hot topic of social research,which has important research value and practical significance for the timely detection and effective positioning of illegal invasion.Traditional security systems are vulnerable to the influence of outside environment.The continuous distributed optical fiber intrusion location system based on optical fiber sensing technology has the advantages of high sensitivity,anti-interference,no dead angle and location of disturbance signals.The sensor system is affected by noise and external environment,which leads to high false alarm rate and low positioning accuracy.In order to recognize intrusion efficiently and locate it quickly and accurately in complex environments,this paper focuses on the identification and localization of vibration signals by distributed optical fiber intrusion location system.The two-stage vibration pattern recognition method and the quadratic correlation algorithm are used to locate the intrusion signal in noisy environment.The experimental results show that the algorithm reduces the false alarm rate of the system and improves the positioning accuracy.The main work of this thesis includes:(1)On the basis of elaborating the principle of double Mach-Zehnder interferometric optical fiber sensor,the feature extraction and recognition algorithm of vibration signal is studied.After pretreatment of the collected signal,the effective group is extracted by setting the threshold of short-term energy and short-term average zero-crossing rate,and the first-level prediction in time domain is completed.Then,the kurtosis and time-frequency entropy are used as time-domain and time-frequency-domain features to form feature vectors for the extracted effective group signals,and BP neural network algorithm is used to recognize the climbing,knocking and raining signals in a variety of ways to complete the second-level recognition.The effectiveness and reliability of the algorithm are proved by a large number of experiments on the monitoring data collected in the field.(2)Aiming at the influence of noise on positioning accuracy in optical fiber sensing signals,cross-correlation algorithm,quadratic correlation algorithm and generalized quadratic correlation algorithm are used to compare the positioning accuracy of climbing signals in different signal-to-noise ratio environments.The experimental results show that the generalized quadratic correlation algorithm has the highest positioning accuracy in low signal-to-noise ratio environments,and can accurately locate human destructive behavior.The accuracy and real-time positioning of security system in miscellaneous environment provides algorithm support.(3)Based on LabVIEW and MATLAB software platform,the programming of modules such as short-time energy and short-time zero-crossing rate,empirical mode decomposition and Hilbert transform,BP neural network,generalized quadratic correlation and positioning are realized by using modular design idea.The design of distributed optical fiber sensor positioning system is completed.The feasibility of the system is proved by experiments.