Fiber Optic Security Intrusion Detection And Positioning Method Of Vibration Signal

Optical fiber sensor is a kind of passive sensors, contains the advantages of corrosion resistance, small volume, light weight, high stability and so on. Optical fiber sensor is widely used in perimeter security system. Because of the nonlinear, non-stationary and intermittent signal in fiber security perimeter system, it is becoming a one of the hotspots, that we find a method to improve the recognition rate of intrusion vibration signal and to further improve the localization accuracy. This paper, aiming at the characteristic of non-stationary vibration signal of intrusion and intermittent, using of chaos theory, studies the signal frame length determination method; using of the characteristics of time domain and frequency domain, gives the method of identification and location of the intrusion signal. Simulation results show that the vibration signal recognition of intrusion and positioning method which is given in this paper has the advantages of high recognition rate, accurate positioning.This paper mainly do the following several aspects of research work:1, This paper puts forward the frame method using saturated embedding dimension to get the minimum length of invasion vibration signal. Using Chaos theory it calculate and analyses the dynamic characteristics of vibration signal, based on the method of saturated embedding dimension to determine the minimum frame length of the time series. By using these methods this paper overcome the nonlinearity and determine the effect of intermittent signal frame length, reduces the computational complexity.2, This paper gives a two invasion of vibration signal identification method based on time-frequency domain. First by using of short-time energy and short-time average zero crossing rate discrimination to certain the vibration signal, second by using of the wavelet coefficients energy distribution characteristics of vibration signal of information to identify the intrusion. The simulation results show that the method can effectively reduce the invasion of vibration signal leakage identification rate and the error rate.3, This paper researches on the positioning method of intrusion of vibration signal. Because of the vibration signal containing the noise, influencing the positioning accuracy, so this paper use a filtering method based on Bayesian adaptive threshold in the domain of wavelet, to denoising the intrusion vibration signal, and than translates reconstruction signal to frequency domain mapping to identify trap points,and last etermine the position of the intrusion signal. The simulation results show that the proposed method can effectively improve the accuracy of intrusion of vibration signal.