摘要
针对实际生产现场的离心式压缩机转子振动信号伴随强噪声干扰,难以进行转子运行状态监测和评价的问题,提出基于Hankel矩阵的主分量分析(Principal component analysis,简称PCA)降噪方法。该方法在以转子振动信号建立Hankel矩阵的基础上,采用主分量分析进行噪声压缩与信号重构。仿真正弦信号和离心式压缩机转子振动信号的分析结果表明,该方法能够实现转子振动信号的高效降噪,得到的降噪结果具有清晰恢复真实信号的能力,有助于生产现场的实时应用,保障生产的顺利进行。
Aiming at the difficulties in monitoring and evaluating the running conditions of rotors with strong noise interference in the centrifugal compressor rotor vibration signal in the actual production field,this paper puts forward the Principal component analysis(PCA)noise reduction method based on the Hankel matrix.Principal component analysis is applied to suppress noise and reconstruct signals by establishing Hankel matrix with rotor vibration signal.The analysis results of simulation sine signal and the signal of the centrifugal compressor rotor vibration indicate that the method can reduce the noise in rotor vibration signals efficiently.And the noise reduction results has the ability to restore the real signal clearly,which helps to the real-time application in production field and ensures the smooth production operation.
出处
《石油化工设备技术》
CAS
2015年第3期40-43,7,共4页
Petrochemical Equipment Technology
关键词
主分量分析
HANKEL矩阵
离心式压缩机
转子
Principal component analysis(PCA)
Hankel matrix
Centrifugal compressor
Rotor
作者简介
钱广华，1982年毕业于华东石油学院化工机械专业．工学学士，在化工部从事设备管理工作，已发表论文多篇，高级工程师。EmaU：qianguanghua．tjsh@sinopec．com。
