经验模态分解法在光纤光栅围栏报警系统中的应用  

Empirical Mode Decomposition in FBG Fence Alarm System

作　　者：蒋祺[1] 饶云江[1,2] 

机构地区：[1]重庆大学光电工程学院光电技术及系统教育部重点实验室,重庆400044 [2]电子科技大学通信与信息工程学院宽带光纤传输及通信网络技术教育部重点实验室,四川成都610054

出　　处：《光学与光电技术》2009年第3期16-20,共5页Optics & Optoelectronic Technology

基　　金：国家自然科学基金(60537040);教育部博士点基金(A0901010120060614017)资助项目

摘　　要：为了解决基于FBG传感器的光纤围栏报警系统中多个FBG传感器在同时受扰时,定位报警信号难的问题,实现对报警信号的有效识别和判断,提出了一种基于经验模态分解(EMD)和小波包特征熵算法的分析方法。利用经验模态分解法对于信号的突变性敏感和有效保留以及其特有的自适应性分解特性,首先对报警信号进行经验模态分解,再结合小波包分解,得到小波包系数提取其信号的能量分布,再做归一化得到信号的能量分布特征向量,进而运用相关性分析实现对报警信号的识别和判断。通过建立实验模型,对采集到的报警信号做了分析,证明了该方法对于解决光纤围栏报警系统中FBG传感器的级联判断报警信号的有效性。In order to solve the problem of located alarm signal in the fiber-based FBG sensor fence alarm system when more than two FBG sensors are disturbed at the same time, and to achieve the goal of effective identification and judgment of alarm signal, a method based on empirical mode decomposition （EMD） and entropy algorithm of wavelet packet characteristics is proposed. First, the alarm signal is decomposed by EMD, which is sensitive to sudden change and effective mutation of the reservation and unique adaptability of decomposition. Then, combing with wavelet packet, the energy distribution and that of vectors of the signal are received. Furthermore, the alarm signal identification and judging are realized by related analysis. Through the establishment of the experimental model, the alarm signal is acquired to be analyzed, and the effectiveness of the method is proved.

关 键 词：经验模态分解 小波包 FBG传感器 光纤围栏报警系统 

分 类 号：TN929.11[电子电信—通信与信息系统]