基于HHT及LCS的轨道电路传输变化识别探讨

在轨道电路的实际使用中,受设备工作性能、使用环境的影响,轨道电路传输特性可能发生变化,导致轨道电路传输曲线产生变形。本文以轨道电路区段为单位,先提取主轨道电路传输曲线,再利用希尔伯特变换(HHT)方法提取趋势分量,并利用最长公共子序列(LCS)算法进行趋势分量相似性对比,探测该区段轨道电路感应电压曲线是否存在变形,用来判断轨道电路传输特性是否发生变化。仿真结果表明,该方法能准确识别轨道电路传输特性变化现象,可为轨道电路的故障识别、故障预测、健康管理研究提供支持。

Exploration on Recognizing Changes in Track Circuit's Transmission Characteristics Based on HHT and LCS

In the actual operation of track circuit,the transmission characteristics of the track circuit may change due to the influence of equipment performance or application environment,resulting in the distortion of its transmission curve.So,the transmission curve of the main track is extracted firstly.Then the trend component is obtained by Hilbert-Huang transform(HHT).Finally,the longest common subsequence(LCS)algorithm is used to compare the similarity of the trend components to detect the deformation of transmission curve,so as to judge whether the transmission characteristics of track circuit have changed.The simulation results show that changes in the transmission characteristics of track circuit can be identified accurately by the proposed method,which can provide support for fault identification,fault prediction and health management researches of track circuit.