基于EMD-Hilbert包络谱分析的涡轮泵轴承故障特征识别
臧东情,秦雷,何伟锋,张栋,刘鑫,刘瑞欣

(北京航天动力研究所,北京 100076)

火箭发动机; 涡轮泵轴承; 故障特征提取; EMD; Hilbert包络解调

Fault feature identification of turbopump bearings based on EMD-Hilbert envelope spectrum analysis
ZANG Dongqing, QIN Lei, HE Weifeng, ZHANG Dong, LIU Xin, LIU Ruixin

(Beijing Aerospace Propulsion Institute, Beijing 100076, China)

rocket engine; turbopump bearings; fault feature extraction; EMD; Hilbert envelope demodulation

备注

在火箭发动机涡轮泵高速轴承试验中,轴承的可靠性对保证试验的成功至关重要。针对涡轮泵轴承故障特征难以从原始信号中提取的问题,基于EMD-Hilbert包络解调分析方法,对涡轮泵高速轴承故障特征进行识别。采用EMD方法对原始信号进行自适应分解,获得若干个IMF分量; 基于相关性指标最大原则筛选IMF分量进行信号重构; 对重构信号进行Hilbert包络解调分析,提取出故障轴承的特征。以某型号涡轮泵低温高速轴承试验的真实故障数据验证本方法的有效性,数据记录了试验装置在阶梯式升速全过程中保持架故障的振动加速度信号。分析结果表明,基于EMD-Hilbert包络解调分析方法能够提高信噪比,最大程度保留保持架故障信息的周期性冲击成分,并能有效提取保持架故障频率、故障倍频及各种调制频率成分,实现对涡轮泵高速轴承故障的有效识别。为深度解析轴承保持架故障情况,结合系统11种运行状态,提出了一种轴承渐进劣化全过程的解析方法,确定出轴承故障早期人为干预的具体时刻。
In the low temperature and high speed bearing test of the rocket engine turbopump, the reliability of the bearing is very important to ensure the success of the test. Aiming at the problem that the early fault features of turbopump high-speed bearings are difficult to extract from the original signal, based on the EMD-Hilbert envelope demodulation analysis method, the early fault features of turbopump high-speed bearings are identified. The original signal is adaptively decomposed by the EMD method to obtain several IMF components, and the IMF components are screened based on the principle of maximum correlation index to reconstruct the signal. The Hilbert envelope demodulation analysis is performed on the reconstructed signal to extract the early features of the faulty bearing. The validity of the method is verified by the real fault data of a certain type of turbopump at low-temperature and high-speed bearing test, and the vibration acceleration data of the cage failure in the test device during the whole process of stepped acceleration are recorded. The analysis results show that the EMD-Hilbert envelope demodulation analysis method can improve the signal-to-noise ratio, retain the periodic impact component of the cage fault information to maximum extent, and effectively extract the cage fault frequency, fault frequency double and various modulation frequency components, so as to achieve the effective identification of early faults of turbopump high-speed bearings. In order to deeply analyze the fault of bearing cage, combined with the 11 operating states of the engine system, an analytical method of the whole process of progressive deterioration of bearing is proposed in this paper, and the specific moment of early human intervention of bearing failure is determined.
·