分布式光纤振动传感器的自适应时延估计

Adaptive time delay estimation of distributed fiber optic perturbation sensor

针对传统的安全防卫系统存在监测距离较短、抗电磁干扰能力弱、维护成本高等缺点,设计了一种新型的分布式光纤振动传感器.传感器由双向马赫-泽德(M-Z,Mach-Zehnder)干涉仪构成全光纤的光路组成,通过光纤敏感入侵者的振动来进行传感和定位.实验测得传感器的信号为低通带限信号.根据传感器的信号特点,研究了基于最小均方(LMS,LeastMean Square)算法的自适应时延估计方法,给出了自适应算法的性能分析结果.建立了实验系统,现场试验结果表明自适应算法可实时地获取振动的位置,平均误差小于200m,已成功用于多个地区的安防报警系统.

Considering the defects of shorter monitoring distance, the sensitiveness to electromagnetic in- terference and higher cost of maintenance in traditional security guard system, a novel distributed fiber optic perturbation sensor was proposed. Composed of the dual M-Z （Mach-Zehnder） interferometers of all fiber instruments, the sensor proposed can locate the position of the intruder and send out the alarm by the sensitive- ness to perturbation of fibers. The experiments show that the signal of intruder＇s perturbation is band-limited. According to the perturbation signal characteristics of the sensor, the self adaptive time delay estimation using the LMS （least mean square） algorithm was studied, and the performance of the self adaptive algorithm was provided. The experiment results show that the self adaptive algorithm can be applicable for locating on real time with the locating error less than 200 meters. The sensor designed has been successfully applied in security guard system in several areas.