基于弗雷歇距离的道岔故障诊断方法

Method of Turnout Fault Diagnosis Based on Fréchet Distance

摘要:

对现场获得的各种铁路道岔故障曲线进行分析，获得几类典型故障的参考曲线，并作为诊断的模板曲线.根据基于弗雷歇距离定义的相似度函数，计算待识别曲线与各模板曲线的相似度.将相似度最大的模板曲线所对应的故障模式作为待识别曲线的诊断结果输出.实验表明，该方法无需大量的训练样本，也不需要对动作电流曲线进行分段特征提取,诊断方法准确率高、速度快、适应性强.

Abstract:

The consequences of frequent occurrence of turnout fault on the railway, are often very serious, which requires a more complete operation of the state monitoring mechanism and intelligent fault diagnosis methods to improve the safety and reliability of turnout equipment. First, according to the curves of various turnout faults obtained on site, the reference curves of several typical faults were selected as a template curve of diagnosis. Then, according to the similarity based on the Fréchet distance，the similarity degree between the test curve and the template curve was calculated, and the failure mode corresponding to the template curve with the largest similarity was output as the diagnosis result of the test curve. The experiments show that this method does not need a lot of training samples, and it is not necessary to extract the segmentation feature of the action current curve. The diagnostic method has a high accuracy, a high speed, and a strong adaptability.