基于阶比分析与AR谱估计的发动机故障特征提取  

Feature Extraction of Engine Faults Based on Order Ratio Analysis and AR Spectral Estimation

作　　者：乔龙[1] 张玲玲[2,3] 陈祥龙[1] 肖云魁[3] 贾继德[3] 

机构地区：[1]军事交通学院研究生管理大队,天津300161 [2]军械工程学院火炮工程系,石家庄050003 [3]军事交通学院汽车工程系,天津300161

出　　处：《汽车工程》2013年第5期462-466,共5页Automotive Engineering

摘　　要：提出了一种基于阶比分析和AR谱估计相结合的发动机故障诊断方法。在发动机急加速过程中拾取油底壳振动信号,将等时间间隔采样所得的非稳态信号按等角度间隔重采样,得到角域中的稳态信号,再进行AR谱估计,得到阶比AR谱。最后进行阶比带的谱幅值累加。结果表明,某些阶比带的谱幅累加值分别更明显地反映发动机曲轴轴承和连杆轴承的磨损故障特征。An engine fault diagnosis method based on order ratio analysis and auto regression （AR） spectral estimation is proposed. The vibration signals of oil pan are picked up in the process of engine hard acceleration. The unstable signals sampled at equal time interval are converted into stable signals in angle domain by re-sampling at equal angle interval and then the order-based AR spectra are obtained by AR spectral estimation. Finally an adding-up manipulation of spectra amplitude in each successive order band is proceeded. The results show that some order bands more markedly reflect the wear fault feature of crankshaft bearings and connecting rod bearings respectively.

关 键 词：发动机 阶比分析 AR谱估计 故障特征 

分 类 号：TH165.3[机械工程—机械制造及自动化] TK407[动力工程及工程热物理—动力机械及工程]