基于 E-S 方法的齿轮系故障信号辨识技术研究

针对齿轮系早期故障信号弱，特征不明显的问题，提出一种基于经验模态分解（EMD）与奇异谱分析（SSA）相结合的故障特征信号辨识方法。通过对信号分解和特征筛选，重组轨道矩阵，实现对不同特征信息的提取和识别。试验表明该方法不仅有效提取出了系统的故障特征信息，而且实现了对系统状态和微弱故障信息的有效识别。

Research on Signal Identification Technology for Failure Gearbox with E-S Method

For the question of signal being weak and feature invisible in early stage of failure gearbox,an identification method of failure feature for the mechanical vibration system based on empirical mode decomposition （EMD） and singular spectrum a-nalysis （SSA） is found. With E-S method, the signal is decomposed, the characteristic is selected, and the attractor orbit matrix can be obtained. And the different diagnostic information of the vibration signal can be extracted with the orbit matrix recombination. It is proved by test that the method can not only effectively extract the characteristics of the system, but also ef-fectively recognize and identify the weak failure characteristics and failure status for the gearbox system.