基于奇异值分解的航空发动机转子碰摩故障特征提取方法

The extraction of the fault features from aero-engine rotor contact-rubbing based on singular value decomposition

提出了利用奇异值分解(SVD)提取航空发动机转子碰摩故障特征信号的方法,通过数值仿真得到,奇异值分解方法可以非常有效地将各种不同特征信号完全提取分离,论证了该方法的可行性。然后以某型国产航空发动机在试车台开车过程中某稳定状态的转子振动测量数据为基础,利用奇异值分解的优选差分谱理论对振动信号进行了降噪处理;根据傅里叶变换频谱图中各频率点峰值与奇异值分解差分谱峰值序列相对应的特点,对降噪后振动信号进行了特征提取,实现了转子系统的碰摩故障特征信号的提取。实际结果表明,该方法能够有效地诊断转子系统碰摩故障及提取相应的故障特征信号。

A singular value decomposition(SVD)method is proposed to extract the fault feature of rotor contact-rubbing of aero-engine.Firstly,the feasibility of this method is demonstrated through the numerical simulation.And then the method of SVD is introduced to the noise reduction of vibration signal,which was based on some stable state of the test bed of some certain type of domestic aero-engine.In order to extract the fault characteristic signals of the rubbing faults of rotor system,according to the characteristics of the corresponding peak sequence between Fourier transform spectrum of each frequency point and singular value decomposition difference spectrum,the feature is extracted from the vibration signal after the noise reduction.Practical results show that the method proposed can effectively diagnose the rubbing faults of rotor system and extract the fault characteristic signals.