【摘要】 研究了基于短时AR分析、小波多分辨率分析和小波包分析的故障特征提取和识别方法,分析了柴油机气缸盖振动信号特征提取方法。得出了两条重要结论:基于短时AR分析的柴油机气缸盖振动信号整循环特征提取方法特别适合于短序列数据的分析;利用小波多分辨率分析和小波包分析以及Kllback-Leibler信息量最小,对柴油机表面振动信号进行分解与分析,确定各故障状态的特征频带,进而可用频带的时间序列的时序模型作为特征矢量,实现对柴油机运行状态故障的诊断。更多还原

【Abstract】 This paper studied the trouble characteristic extraction and identification methods based on snatch AR analysis,wavelet many resolutions analysis and wavelet packet analysis,and analyzed vibration signal characteristic extraction method for diesel engine cylinder head.So two conclusions were obtained:the vibration signal total cycle characteristic extraction method for diesel engine cylinder head based on snatch AR analysis was fit particularly for the short sequential data analysis;by using wavelet many resolutions and wavelet packet analysis,the surface vibration signal of diesel engine was decomposed and analyzed,the characteristic frequency band of every trouble state was confirmed,then the time sequence model of frequency band time array could be used as the characteristic vector,to realize the diagnosis of diesel engine trouble.更多还原