Design And Disturbance Recognition Of Mach-Zehnder And ?-OTDR Joint Fiber Perimeter System

The optical fiber perimeter system has the advantages of strong electromagnetic interference resistance,long sensing distance,high sensitivity and full distribution.It has important application prospects in the perimeter security of civil aviation airports.The system has a complicated working environment and various types of disturbance intrusions.Traditional optical fiber perimeter systems also have problems such as high false alarm rates and inability to simultaneously have excellent identification and positioning performance,which cannot meet the airport's performance requirements for the perimeter system.In view of this problem,this paper designs Mach-Zehnder and ?-OTDR joint fiber perimeter system,so that the fiber optic perimeter system has excellent recognition and positioning performance at the same time.And further study the signal processing method of the joint optical fiber perimeter system to improve the application performance of joint optical fiber perimeter system.The thesis has done the following research work:1.Based on the analysis of the advantages and disadvantages of the traditional optical fiber perimeter system,the overall design scheme of the Mach-Zehnder and ?-OTDR joint optical fiber perimeter system is proposed and the theoretical model of the joint optical fiber perimeter system is constructed.According to the performance requirements of the airport perimeter for the optical fiber perimeter system,the selection of key components of the joint optical fiber perimeter system and the construction of the joint optical fiber perimeter system are completed.2.Aiming at the problem that the signal-to-noise ratio of the ?-OTDR subsystem in the joint fiber perimeter system is too low,a denoising method based on improved variational mode decomposition is studied.Experimental verification was performed using the measured signals of the joint fiber perimeter system.The results show that the method can effectively remove noise.Compared with traditional empirical mode decomposition-independent component analysis and simulated annealing-variational mode decomposition method,the signal-to-noise ratio is improved by 4dB,which effectively improves the joint fiber perimeter system signal-to-noise ratio.3.Aiming at the problems of the disturbance information extraction of disturbance signals,the relatively small number of recognition disturbance types and the recognition accuracy is relatively low of the Mach-Zehnder subsystem in the joint optical fiber perimeter system,a method for extracting perturbation information and pattern recognition based on local feature scale decomposition and improved probabilistic neural network is studied.Experimental verification was performed using the measured signals of the joint fiber perimeter system.The results show that the method can efficiently and accurately identify six types of disturbance signals.The average recognition rate is 97.17% and the recognition time is 0.78 s,which effectively improves the identification and application performance of the joint optical fiber perimeter system are presented.