基于小波奇异性检测的发动机故障诊断方法研究  

A Study on Engine Fault Diagnosis Method Based on Wavelet Singularity Detection

作　　者：肖云魁[1] 乔龙[1] 张玲玲[1,2] 赵慧敏[1] 杨青乐 

机构地区：[1]军事交通学院汽车工程系,天津300161 [2]军械工程学院火炮工程系,石家庄050003

出　　处：《汽车工程》2014年第11期1405-1409,共5页Automotive Engineering

基　　金：总后勤部科研项目(后司[2009]835号)资助

摘　　要：针对通过二进离散小波变换计算出的Lipschitz指数不够精确且不能描述奇异点能量的不足,提出了模极大值点能量法并用于信号的奇异性检测。通过仿真证明采用该方法不仅能准确定位奇异点发生的位置,并且能描述其能量。将该方法应用于发动机曲轴轴承故障诊断时,为了突出局部时间段的信号特征,采用抽区间采样方法抽取特定时间段的信号。试验结果表明,两种方法的结合能有效区分出曲轴轴承的不同技术状况,并得出最佳诊断部位和最佳检测转速。In view of the defects of Lipschitz exponent calculated by discrete dyadic wavelet transform （ in-sufficient accuracy and incapability in expressing energy） , a method of“energy at modulus maxima points” is pro-posed for singularity detection of signals. It is proved by simulation that the method can not only locate the singular points but also describe their energy. For giving prominence of the signal features of local time segment in applying the method to the fault diagnosis of engine crankshaft bearings, the scheme of extract time interval sampling is a-dopted to extract the signals in specific time segment. Test results show that the combination of“energy at modulus maxima points” method with extract time interval sampling scheme can effectively distinguish different technical con-ditions of crankshaft bearings with the best location for diagnosis and the best rotating speed for detection obtained.

关 键 词：发动机 故障诊断 二进离散小波变换 模极大值点能量 奇异性检测 抽区间采样 

分 类 号：U472.9[机械工程—车辆工程]