基于IK-means和CHMM的铁路接触网性能退化评估  

Assessment of Performance Degradation of Railway Catenary Based on IK-means and CHMM

作　　者：侯绪永 HOU Xuyong(China Railway Construction Electrification Bureau Group No.1 Engineering Co.,Ltd.,Luoyang,Henan 471013,China)

机构地区：[1]中国铁建电气化局集团第一工程有限公司,河南洛阳471013

出　　处：《机车电传动》2021年第2期140-145,共6页Electric Drive for Locomotives

基　　金：中国铁路总公司科技研究开发计划重点课题(2017X011-C)。

摘　　要：了解铁路接触网的退化性能水平有助于保障机车的安全运行。为了提高传统K-means算法聚类的合理性和准确性,引入了密度值因子来提高数据聚类精度,称为IK-means算法。采用IK-means算法对高速铁路接触网进行模式划分,将铁路接触网的退化状态划分成正常状态、轻度退化、中度退化、重度退化和失效状态共5种状态,再采用隐马尔科夫(Hidden Markov Model, HMM)算法对铁路接触网的性能退化状态进行评估。仿真测试结果表明,评估精度平均高于92.67%,验证了所提方法的可靠性。Understanding the degradation level of railway catenary is helpful to ensure the safe operation of locomotives. In order to improve the rationality and accuracy of the traditional K-means algorithm, the density value factor was introduced to improve the precision of data clustering, which called IK-means. The IK-means algorithm was used to divide high-speed railway catenary mode, which divided a railway catenary state of degradation into five types, including normal, mild degradation, moderate degradation, and severe degradation and failure state. The condition of the railway catenary degradation was assessed through CHMM. Simulation test results showed that the assessment accuracy was higher than 92.67%, which verified the reliability of the proposed method in this paper.

关 键 词：高速铁路 接触网 性能退化 IK-means算法 CHMM算法 评估 

分 类 号：U225[交通运输工程—道路与铁道工程]