基于IMF和粗糙度特征的发动机振动信号分析

摘要

航空发动机振动信号为典型的非平稳信号,包含了多种振源振动信息和大量的噪声分量。通过对航空发动机振动信号进行Hilbert-Huang变换,将复杂信号分解为代表不同物理意义的单分量固有模态函数（intrinsic mode functions,IMF）,然后对每一个IMF信号提取广义粗糙度特征实现对振动信号的描述。由于各IMF分量的能量百分比大小表征了该分量信号的有效性,使用提取的能量百分比对各分量下的广义粗糙度特征进行加权,最后得到了对发动机振动信号进行描述的能量加权广义粗糙度特征。通过对航空发动机实际试车采集的碰摩振动信号和正常工况下信号的实验分析可以看出,两种情况下信号特征具有明显不同,说明该特征可以有效地对振动信号进行描述。%The vibration signal of the aeroengine is non-stationary and involves multi-frequency components as well as an abundant of noise. The Hilbert-Huang transform was used to decompose the vibration signal into a number of intrinsic mode functions （IMF） which represented the physically meaning of the data. Then, the generalized roughness vectors of the IMF were obtained. Because the energy of the IMF represents the validity of this component in the whole signal, the percentage of energy is used as the weight of the generalized roughness feature. Finally, the energy weighted generalized roughness feature was obtained. The experiments on the aeroengine test data of the rubimpact and the normal conditions demonstrated that these statistic characteristics were distinguished for different conditions.