摘要
针对传统经验模式分解(EMD)确定含噪模式分量缺乏具体评价指标的问题,首先利用互补集合经验模式分解(CEEMD)将故障信号分解,然后利用去趋势波动分析(DFA)计算每一个模式分量对应的标度指数,有用分量和含噪分量通过标度指数的幅值阈值进行区分,最后小波分析用于对识别出的高频含噪分量进行降噪处理,其目的是最大程度地保留高频模式分量中的故障信息,实现信号的自适应降噪。通过对轴承故障信号和数值仿真信号的分析结果表明:提出的方法能够更好地识别和提取轴承的故障特征。
Aiming at the disadvantage of current denoising method using Empirical Mode Decomposition in evaluating the noisy components, a novel fault feature extraction scheme is proposed. The original fault signal is decomposed into a series of IMF by CEEMD. Then, the scaling exponent of each IMFs is obtained through DFA, which the amplitude threshold of each scaling exponent can be used to effectively distinguish the useful signal and the noise signal. Finally, the wavelet denoising was applied to those noisy IMFs, which aims to maximize the retention of the effective components of these IMFs and achieve adaptive noise reduction. The numerical simulation and the measured fault bearing signal are analyzed by the proposed method, which demonstrated that the proposed method can better achieve the identification of rolling bearing fault.
作者
田锐
TIAN Rui(Jingchu University of Technology,Hubei Jingmen 448000,China)
出处
《机械设计与制造》
北大核心
2018年第12期100-104,共5页
Machinery Design & Manufacture
基金
湖北省自然科学基金(2015CFB306)
