• CN:11-2187/TH
  • ISSN:0577-6686

›› 2013, Vol. 49 ›› Issue (8): 1-9.

• 论文 •    下一篇

基于精确信息重构的故障转子系统振动加速度信号积分方法

温广瑞;李杨;廖与禾;何庆   

  1. 西安交通大学陕西省机械产品质量保障与诊断重点实验室;西安陕鼓动力股份有限公司
  • 发布日期:2013-04-20

Faulty Rotor System Vibration Acceleration Signal Integration Method Based on Precise Information Reconstruction

WEN Guangrui;LI Yang;LIAO Yuhe;HE Qing   

  1. Shaanxi Province Key Laboratory of Mechanical Product Quality Assurance and Diagnostics, Xi’an Jiaotong University Xi’an Shaangu Power Co., Ltd
  • Published:2013-04-20

摘要: 在分析总结加速度信号时域积分法与频域积分法优缺点基础上,提出一种基于精确信息重构的故障转子系统振动加速度信号积分方法。该方法利用故障转子系统振动信号由转频、倍频及分倍频分量构成为主的特点对加速度信号进行快速傅里叶变换(Fast Fourier transform, FFT),通过特征频率分量提取并将所提取的分量的幅值与预设的阈值进行比较;幅值低于阈值的分量认为是噪声分量,予以剔除,高于阈值的分量保留并进行精确频谱校正;根据校正后各个频率分量的频率、幅值和相位积分重构出相应的速度信号和位移信号。精确信息重构方法在去除宽频噪声和保留有用特征信息方面有明显的优势。最后通过仿真分析和试验验证,结果表明该方法相对于传统的时域和频域积分具有更高的精度和优越性。

关键词: 精确重构, 快速傅里叶变换, 频谱校正, 误差评价, 阈值

Abstract: A faulty rotor system vibration acceleration signal integration method based on precise information reconstruction is proposed in this paper by summarizing the advantages and disadvantages of current integration method both in time domain and frequency domain. This method utilizes the characteristic structure of the faulty rotor system vibration signal, which is mainly composed of the rotational frequency and its superharmonic and/or subharmonic frequency components. The fast Fourier transform is performed to the acceleration signal to get its spectrum, then the characteristic frequency components are searched. Comparing the amplitudes of all the extracted components with a threshold set by SNR. The components, whose amplitudes are below the threshold, will be considered as noise components and will be discarded and only those whose amplitudes are above the threshold will be reserved. The high-accuracy spectrum correction is then made with those components reserved to get the precise information, including frequency, amplitude and phase information, needed for the integration and signal reconstruction. The velocity signal and displacement signal are therefore reconstructed with the precise information of all the reserved components. Precise Information Reconstruction is excellent at removing the broadband noise. The reliability and effectiveness of the proposed method are testified by the simulation analysis and the application examples.

Key words: Error evaluation, Fast Fourier transform, Precise information reconstruction, Spectrum correction, Threshold

中图分类号: