从海缆历史温度数据中分析出海缆的老化或者故障位置是工程应用中急需解决的问题。针对上述问题, 采用EMD(经验模态分解)算法提取出海缆每一处历史温度数据的时频特征, 再通过矩阵相似度来衡量不同位置点海缆时频特征的相似度大小, 最后通过分析对比相似度的异常来确定海缆老化或故障的位置。实例验证结果表明, 问题海缆位置点与正常海缆位置点之间的相似度远小于正常海缆位置点之间的相似度, 因而能预测出所有问题海缆的位置点。

Analyzing from the data of historical temperature to decide the situation of bugs or aging of submarine cable is a critical issue in practical situation. To solve these problems, Empirical Mode Decomposition (EMD) is used toextract the time-frequency characteristics of historical temperature of the submarine cable. Then, the similarity of the frequency characteristics of the submarine cable is examined via matrix calculation at different positions. Finally, the position of the problematical marine cable will be located through the analysis of the abnormality of the similarity.The experimental result shows that the similarities between the positions of problematical submarine cable are bounded to less than the regular one. Therefore, the position of the problematical marine cable can be accurately located.