基于区域划分和边缘检测的Φ-OTDR定位方法

Φ-OTDR Localization Method Based on Region Segmentation and Edge Detection

针对相位敏感光时域反射(Φ-OTDR)技术在扰动定位中数据处理量过大的问题,提出了一种将区域划分和边缘检测相结合的定位方法。通过对光纤全链路进行区域划分,实现对扰动所在区域的快速粗略定位,在得到的扰动目标区域内引入基于Sobel算子的边缘检测算法,进一步实现对发生扰动位置的精细定位。实验中采用相干探测结构的Φ-OTDR系统采集数据,对5 km传感光纤上的三种扰动事件进行定位,其结果表明,基于区域划分的粗略定位方法增强了系统在噪声中的扰动定位能力,对每个样本的平均处理时间仅为0.046 s,较传统方法减少一个数量级。此外,将所提出的结合方法与单独使用边缘检测的算法进行比较,证明所提方法能够在保证定位精度的同时提高定位实时性,且定位误差小于2.84 m,在对外部扰动的实时定位中具有潜在的应用价值。

In response to the large amount of data processing in phase-sensitive optical time-domain reflection(Φ-OTDR)technology for disturbance localization,a positioning method combining region segmentation and edge detection is proposed.The fast coarse localization of the disturbance occurrence region is achieved by dividing the optical fiber into segments,followed by fine localization using an edge detection algorithm based on the Sobel operator within the disturbance region.A heterodyne coherentΦ-OTDR system is employed for data acquisition in the experiments to localize three types of disturbance events on a 5 km sensing fiber.The results show that the coarse positioning method based on region segmentation enhances the system's ability to locate disturbances in noise.The average processing time for each sample is only 0.046 s,which is an order of magnitude lower than that of the traditional method.In addition,the proposed combined method improves real-time localization over using edge detection algorithm alone,while maintaining localization accuracy.The localization error is less than 2.84 m,indicating potential for application in accurately locating external disturbances in real-time systems.