小波包分解与共振解调分析在发动机配气机构磨损故障特征提取中的应用  

Application of Wavelet Packet Decomposition and Resonance Demodulation Analysis to Feature Extraction of Engine Valve-train Wear Fault

作　　者：喻菲菲[1] 杜灿谊[1] 曾祥坤[1] 杜宝雷[2] 

机构地区：[1]广东技术师范学院机电学院,广东广州510635 [2]华南理工大学机械与汽车工程学院,广东广州510641

出　　处：《车用发动机》2015年第1期87-92,共6页Vehicle Engine

基　　金：国家自然科学基金(51305085)

摘　　要：以常见的机械磨损故障———气门间隙过大故障为例进行了试验研究，提取缸盖振动加速度信号进行分析，发现故障状态的时域信号有明显低频周期性冲击，但在频谱的低频区间未现冲击频率；同时，在故障状态频谱中，3000～4500 Hz范围的高频段振动能量有显著增加。通过小波包分解方法对信号分解至该故障特征频段，再进行希尔伯特解调分析，解调谱现显著的对应低频冲击的频率成分，可作为故障识别特征。分析结果表明，气门间隙过大故障造成的冲击引起缸盖或其部件共振调制现象，综合运用上述时域、频谱和共振解调分析，可对配气机构磨损故障进行故障特征提取，从而为准确诊断故障提供依据。For the mechanical wear fault of large valve clearance ,the acceleration signal of cylinder head vibration was collect‐ed by the experiment .It was found that there was low frequency periodic impact in time domain and was not impact frequency in frequency domain in fault state .Meanwhile ,the vibration energy in 3 000‐4 500 Hz increased evidently in the spectrum of fault state .The signal was decomposed to the fault characteristic frequency band by the wavelet packet decomposition and was demodulated by Hilbert demodulation analysis .The frequency components of low frequency impacts in demodulation spectrum were found ,which could be used as the feature of fault diagnosis .The results show that the impact of large valve clearance fi‐nally leads to the resonant modulation phenomenon of cylinder head or other components .Therefore ,the feature of valve‐train wear fault can be extracted by the time domain ,frequency spectrum and resonance demodulation analysis ,which provides the basis for accurate fault diagnosis .

关 键 词：小波包分解 共振解调 配气机构 气门间隙 磨损 故障特征 

分 类 号：TH113.1[机械工程—机械设计及理论]