铁道科学与工程学报2018，Vol.15Issue(6)：1399-1405,7.
基于Copula函数的高速列车信号联合特征提取
Extracting joint feature of signals of high-speed train using Copula
摘要
Abstract
Aiming at mechanical vibration signals of different channels of high-speed train bogie, an approach using Copula function to extract joint feature of signals among different channels was proposed in the paper. The marginal distribution functions of the signals were fitted by generalized Gaussian distribution. The joint distribution function of the marginal distribution functions was computed by Gaussian Copula function. The coefficients of marginal distribution functions and joint density function were extracted as the features. Vibration signals of a high-speed train bogie were obtained under four typical working conditions including normal condition, yaw damper fault, air spring fault, and lateral damper fault. The features were extracted using the proposed approach, and the working conditions were classified with the support vector machine. The average recognition rate was above 97％, which verified the effectiveness of the proposed feature extraction method.关键词
高速列车/通道间信号/Copula函数/联合特征提取Key words
high-speed train/signals among different channels/Copula function/joint feature extraction分类
信息技术与安全科学引用本文复制引用
颜云华,金炜东..基于Copula函数的高速列车信号联合特征提取[J].铁道科学与工程学报,2018,15(6):1399-1405,7.基金项目
国家自然科学基金重点资助项目(61134002) （61134002）
