基于S变换和图像纹理信息的轴承故障智能诊断方法

旋转机械设备发生故障时产生的振动信号具有非平稳、非线性的特点,而传统的基于傅里叶分析的方法不仅不能有效诊断故障,同时需要技术人员具备大量的专业知识,因此提出了基于时频图像纹理信息的智能故障诊断方法。分析几种时频分析方法的优缺,在此基础上对振动信号采取S变换构建时频图像,并利用图像的灰度-梯度共生矩阵提取纹理特征,最后采用支持向量机实现多类故障的诊断。实验验证了方法的有效性。

An intelligent bearing fault diagnosis method based on S transform and texture image

Traditionally, Fourier transform is used to analyse faulty signals ,which demand technicians specific knowledge to diagnose. Since vibration signals from faulty rotating machinery are unstable and nonlinear , this paper presents an intelligent diagnosis method based on time-frequency image . The advantages and disadvantages of several time-frequency analysis method are compared. On the basis of the result , S transform is utilized to obtain images ,and gray level-gradient co-occurrence matrix is calculated to extract texture feature from images. Finally, we adopt support vector machine to realize multi-classification. And, Experimental result demonstrates the effectiveness of the method.