基于时频域模型的噪声故障诊断

为了避免传统的基于振动信号的内燃机主轴承磨损故障诊断中安装传感器以及提取故障特征频率的麻烦,采用一种基于内燃机工作噪声信号和时频域分析的方法。首先讨论了对内燃机噪声信号进行小波包络谱分析,得到可以判断主轴承磨损故障的特征频率。然后,进一步阐述了采用噪声信号小波包分解,可得到包含更多故障信息时-频分布图。基于此,运用图像处理技术建立基于图像匹配的内燃机主轴承诊断模型。结果表明此方法简单有效,充分利用了故障信息。

Noise Fault Diagnosis Based on Time-frequency Domain Model

In order to avoid the difficulty of installing vibration sensors and extracting characteristic frequency vectors in the traditional vibration-based abrasion faults diagnosis technique for main bearing of diesel engine, a new approach based on the time-frequency domain analysis of noise signal of diesel engine is applied to fault diagnosis of main bearing . The wavelet envelope spectrum analysis of noise signals of diesel, from which the characteristic frequency vector representing the gap abrasion condition of main bearing can be deduced, is firstly presented. Then, the fault diagnosis method with the wavelet packet decomposition of noise signal is illustrated in detail. Meanwithle, the standard time-frequency distribution images of all fault conditions, containing the gap abrasion information of main bearing, can be defined. Based on that, a fault diagnosis model for main bearing with images mathing is set up. Through comparing the values of Enclid Distance between standard fault images and the diagnosed image, the gap condition can be recognoized. The result shows that the method is simple and effective, and makes the best use of fault information.