摘要
针对现有振动信号降噪方法中经验模态分解存在模态混叠、独立分量分析要求采集的振动信号数不少于源信号数等问题,提出了一种基于集合经验模态分解(EEMD)和快速独立分量分析(FastICA)的矿用带式输送机驱动滚筒轴承振动信号降噪方法。首先,通过EEMD算法对采集的振动信号进行分解,得到若干不同尺度的包含故障特征频率的本征模态函数(IMF)分量;然后,基于相关系数对IMF分量进行重构,得到特征信号和虚拟噪声信号,将重构的特征信号和虚拟噪声信号组成输入矩阵,并作为FastICA算法的输入;最后,利用FastICA算法实现信号与噪声分离,达到信号降噪的目的。实验结果验证了该方法的可行性和有效性。
EMD has modal aliasing and ICA requires the number of collected vibration signal should not be less than the number of source signal. In view of the above problems that existed in existing vibration signal denoising method, a vibration signal denoising method for drive roller bearing of mine-used belt conveyor was proposed which was based on EEMD and FastICA. Firstly, collected vibration signal is decomposed by EEMD algorithm, so as to obtain several IMF components at different scales that containing faults characteristic frequency. Then, the IMF components are reconstructed based on correlation coefficient to obtain characteristic signal and virtual noise signal, which are formed into input matrix as input of FastICA algorithm. Finally, FastICA algorithm is used to separate signal and noise to achieve purpose of signal denoising. The experimental results verify feasibility and effectiveness of the method.
作者
杨祥
田慕琴
李璐
宋建成
张林锋
吴静坤
YANG Xiang;TIAN Muqin;LI Lu;SONG Jiancheng;ZHANG Linfeng;WU Jingkun(National & Provincial Joint Engineering Laboratory of Mining Intelligent Electrical Apparatus Technology, Taiyuan University of Technology, Taiyuan 030024, China;Shanxi Key Laboratory of Mining Electrical Equipment and Intelligent Control, Taiyuan University of Technology, Taiyuan 030024, China)
出处
《工矿自动化》
北大核心
2019年第3期66-70,共5页
Journal Of Mine Automation
基金
山西省科技重大专项计划基金资助项目(20131101029)
关键词
矿用带式输送机
驱动滚筒轴承
滚筒轴承故障诊断
振动信号降噪
集合经验模态分解
快速独立分量分析
mine-used belt conveyor
drive roller bearing
fault diagnosis of roller bearing
vibration signal denoising
ensemble empirical mode decomposition
fast independent component analysis
