基于SPWVD识别的滚动轴承智能检测方法

Intelligent fault diagnosis methods of rolling bearing based on SPWVD and AIN

为了探索基于振动谱图像模式识别的智能故障检测方法,以滚动轴承为对象,提出了用SPWVD分布来表征振动信号时频分布特性,利用SPWVD图像的GLCM及其特征统计量来提取故障特征。改进了人工免疫网络分类算法,通过人工免疫网络分类方法对故障样本特征统计量进行学习,形成记忆抗体集,进而对检验抗原进行故障分类识别,在故障特征信号干扰严重的情况下,取得了较BP神经网络好的检测准确率,验证了人工免疫网络良好的适应性。随着智能故障检测技术发展,基于图像模式识别的故障检测方法必将得到推广和应用,验证其在轴承故障监测中的可行性。

Intelligent fault diagnosis methods based on pattern recognition of vibration spectrogram were studied. Firstly,taking rolling bearing as an example,the GLCM was extracted from SPWVD spectrogram and its statistic characteristics were described. Moreover,a modified AIN algorithm was introduced and used in bearing fault diagnosis. Through optimization of fault antigen sample,the memory antibodies sets were formed and classification was processed by the k-nearest neighbor method. A mass of fault sample were analyzed in the algorithm proposed and the results were compared with those obtained by BPNN. The comparison result indicated that the modified AIN algorithm had better classification ability as well as higher diagnosis accuracy. With intelligent fault diagnosis methods developing,methods based on spectrogram identification should be popularized,its practicability was proved through recognition of bearing fault here.