摘要
机械振动或声学信号的处理是状态监测及故障诊断的基础。探讨了经验模式的分解特性,将相干分析应用于判定分解中虚假分量,以避免虚假分量和微弱信号的误判;其次研究了信号各频率能量比例对EMD分解时虚假分量的影响,提出了增加大幅值的高频分量可以减小虚假分量比例的方法;然后基于白噪声的EMD分解特性,提出了以叠加多次有色噪声改善EMD模式混叠的方法,同时仿真分析了其优越性,最后将改进的EMD系统方法应用于旋转机械噪声和排气系统振动信号的特征提取,验证了新方法的可行性。
The vibration and noise signal processing of mechanism is the basis of condition monitoring and fault diagnosis. It discusses the main characteristics of EMD（Empirical mode decomposition）. Using the coherence analysis,the misjudgment of weak signal and false component can be effectively avoided. From the perspective of analyzing frequency-energy of signal components,the influence of false component generating in EMD decomposition is researched and the new conclusion that increasing high frequency component can reduce the ratio of false component is discovered. By studying the EMD decomposing characteristics of white noise,the method of multiple stacking colored noise to improve mode mixing of the EMD is proposed,which simulation effect is obviously improved compared to the classic EEMD（Ensemble Empirical Mode Decomposition）. The system method improved EMD is applied to extract signal features from rotating machinery and vibration from exhaust system,and the new method is verified to be feasible and effective.
出处
《机械设计与制造》
北大核心
2016年第2期98-102,共5页
Machinery Design & Manufacture
基金
国家自然科学基金项目(51405221)
南京工程学院校级科研基金项目(YKJ201334)
江苏省自然科学基金(BK20130746)
关键词
机械信号
经验模式分解
虚假分量
模式混叠
Mechanism Signal
Empirical Mode Decomposition
False Component
Mode Mixing
