摘要
提出一种研究旋转机械瞬态信号的分析方法。对齿轮箱启动时测得的原始振动信号进行角域重采样,然后对角域里的信号进行EMD(Empirical Mode Decomposition)分解得到多个IMF(Intrinsic Mode Function)分量,最后对各个IMF分量进行了边际谱分析。结果显示:阶次跟踪技术能够有效地避免传统频谱方法所无法解决的"频率模糊"现象,EMD方法能够将稳态信号分解为多个IMF分量,边际谱能够对轴承故障进行准确诊断。将两种方法相结合是对传统的频谱分析法的有力补充,具有很广阔的应用前景。
A new method was brought forward to analyse the instantaneous signals in rotaing mechinery.The original vibration signals at start-up of the gearbox were resampled in angle-domain.The signals in angle-domain were decomposed with EMD,and many IMF components were obtained.Each IMF component was then analyzed with marginal spectrum.The outcome shows that the “frequency smearing” phenomenon,which cannot be solved with the traditional frequency spectrum method,can be effectively avoided by using order tracking technology and the faults in bearing can be exactly diagnosed with marginal spectrum.The approach combining the order tracking analysis with Hilbert-Huang transform marginal spectrum is proved to be a trenchant supplement to traditional spectrum analysis method and will be of wide application foreground.
出处
《振动与冲击》
EI
CSCD
北大核心
2007年第6期1-3,9,共4页
Journal of Vibration and Shock
基金
国家自然科学基金资助项目(50375157)
军械工程学院创新基金项目(CXJJ0504)
