矢谱和HMM理论在风机故障诊断中的应用

The Application of Vector Spectrum and HMM Theory in Blower Fault Diagnosis

传统风机用单一传感器进行故障的特征提取,不能有效地反映其状态。然而,矢谱技术融合了多通道信息,其反应的信息更全面,结合隐马尔可夫模型(HMM)能够对一个时间跨度上的信息进行统计建模和分类的特点,提出了一种新的诊断方法;并在Matlab的环境下给出了诊断的实例,表明该方法可以有效地应用于风机的诊断中。

Traditional blower with single sensor for fault feature extraction can not show the state effectively. The vector spectrum can fuse the information from different channels and reflect more comprehensive information, and combined the characteristics of HMM on the statistical modeling and the classification to a time span of information, a blower fault diagnosis method is studied. An instance of fault diagnosis of blower is presented in Matlab environment, which indicates this method can be applicated effectively in blower fault diagnosis.