基于小波对称极坐标方法的发动机异响故障特征提取研究  

A Reseach on the Feature Extraction of Engine Abnormal Sound Faults Based on Wavelet Symmetric Polar Coordinates Method

作　　者：贾继德[1] 

机构地区：[1]军事交通学院汽车工程系,天津300161

出　　处：《汽车工程》2013年第11期1035-1037,1029,共4页Automotive Engineering

摘　　要：针对发动机异响故障信号包含较强的背景噪声和时频分布图对故障特征表现不直观的问题,提出一种小波对称极坐标方法。用连续小波变换分析振动信号,通过小波系数相关降噪处理提取故障信号特征,然后将信号特征通过对称极坐标方法表示,生成对称点极坐标图;通过识别对称点极坐标图形的显著差异来实现发动机故障诊断。利用仿真信号测试该方法的有效性,并进一步运用该方法来识别发动机常见异响故障,结果表明,该方法可提高信噪比,增强发动机异响故障特征提取的效果。Aiming at the problems of engine abnormal sound fault signal containing strong background noise and the representation of fault feature being not so intuitiuve in time-frequency domain distribution chart,a wavelet symmetric polar coordinate method is proposed. The vibration signals are analyzed by using continuous wavelet transform technique and the fault signal feature is extracted by wavelet coefficients correlation denoising processing,which is then represnted by symmetric polar coordinates and symmetrized dot polar graphs are generated. By identifying the significant difference in those graphs,the fault diagnosis of engine is realized. Simulated signals are used to testify the effectiveness of the method,which is then further utilized to identify the abnormal sound faults of engine.The results show that the method proposed can raise the signal-to-noise ratio and enhance the feature extraction effect of engine abnormal sound faults.

关 键 词：发动机 异响故障 小波降噪 对称极坐标 特征提取 

分 类 号：TK407[动力工程及工程热物理—动力机械及工程]