基于谐波追踪的发动机转子系统早期故障特征提取方法研究

应用谐波追踪方法研究了航空发动机振动问题,对发动机转子系统早期故障表现的高度不平稳信号进行自适应处理。文中在传统Gabor原子库基础上介绍了一种谐波原子,并针对转子异常振动冲击使用了非对称高斯窗函数,对振动信号进行匹配追踪分解,从而准确而有效地获得了转子系统的冲击损伤特征和信息。计算机仿真结果证实了该方法的准确性,最后将该方法应用于转子实验台的冲击信号特征提取中。

Extracting Features of Early Faults in the Rotor System of Aero-engine with Harmonic Matching Pursuit

We apply the matching pursuit algorithm to the adaptive processing of highly unstable signals of early faults in the rotor system of an aero-engine. We present harmonic atoms to extend the Gabor atomic bank and use the asymmetrical Gaussian function to conduct the matching pursuit decomposition of the vibrational signals of the rotor system, thus obtaining the exact and effective features and information on its impact faults. The simulation results verify the accuracy of the above-mentioned method, and it is applied to extracting the features of impulse signals of a rotor rig.