| Optical fiber pre-warning system is a kind of new security means with fiber as the sensor,which can effectively overcome the defects of traditional security system such as low sensitivity,poor electrical insulation and insufficient concealment.In addition,it has advantages of anti-electromagnetic interference,long monitoring distance and real-time.Optical fiber pre-warning system is the new trends in future development of security systems.The optical fiber pre-warning system needs to effectively judge whether the intrusion occurred,accurately report the time and place of the intrusion and identify the types of intrusion so as to help to quickly take targeted measures to reduce the corresponding losses.In order to meet the requirements of the application level and realize the effective detection and recognition,a series of signal processing methods are used to analyze and study the optical fiber signals.An intrusion detection scheme of optical fiber early warning system is proposed and implemented based on DSP.In order to realize the optical fiber pre-warning system intrusion identification in DSP,the task is mainly divided into intrusion detection location and intrusion type identification,and a series of research has been carried out in this paper.In this paper,the constant false alarm false alarm rate(CFAR)detection algorithm is introduced to detect the intrusion signal,and a false alarm rejection algorithm based on the correlation analysis of orthogonal polarization signals is proposed,which can effectively improve the detection effect of the system.In the aspect of intrusion identification,this paper proposes a method of optical fiber intrusion signal recognition based on AMDF mean rate,detection period alarm ratio and frequency domain statistical characteristics.This method can effectively distinguish between picking,mining and vehicle intrusion.Fiber optic intrusion system needs to be implemented in embedded signal processing platform.Due to the limited computing and storage resources of signal processing platform and the high complexity of related algorithms,it is necessary to design a reasonable processing flow of digital signal processor.In this paper,the optical fiber pre-warning system is implemented based on DSP platform.Firstly,the intrusion detection is carried out to determine the time and position of the intrusion,and the result of the detection is taken as the input of the identification to reduce the amount of data needed to be processed in the subsequent identification.In the process of feature extraction,the method of feature extraction is optimized to reduce the computational complexity and ensure the real-time performance of the system.The field test of the optical fiber pre-warning system proves that the intrusion detection algorithm and the DSP implementation scheme proposed in this paper can achieve the corresponding functions efficiently and stably,and meet the expected performance requirements. |
