中文摘要：

光时域反射仪（OTDR）只能测量各取样点处的衰减值,不能自动地判断出事件点.实际光纤线路的信号中常常混有噪声,在故障分析中必须将事件点信息从噪声中提取出来,再精确地判断出事件点的位置.该文在简单分析了OTDR测试原理的基础上,提出了一种基于小波变换的光纤故障分析方法.该算法首先将小波变换后的高频数据分量进行阈值滤波和信号增强处理,在很好地抑制噪声的同时,还能有效地增强突变点信息.然后根据模极大值的位置和特点,对增强后的信号进行检测,以判断事件点的位置和类型.仿真实验结果表明,该算法在较强的噪声干扰的情况下仍然有较高的精度和可靠性,是一种行之有效的方法.

英文摘要：

Otdr can only measure the attenuation value of every sample, but it can＇t judge event point automatically. In real optic fiber circuit, signal will be often mixed with noises. During fault analysis, it＇s necessary to find event information from the noise, and accurately judge position of event points. After simply analyzing otdr test principle, the algorithm of fiber fault analysis based on wavelet transform is proposed. First, the threshold filter and the signal correlation enhancement will be carried out to the high frequency data of wavelet transformation, it can not only denoise commendably, but also strengthen the point efficiently. Then according to position and characteristic of mold maximum value, the strengthen signal will be detected to judge the event position and type. As an effective method, the simulation experiment result indicated that, this algorithm still had high precision and reliability in the strong noise jamming situation.