论文部分内容阅读
提出一种集合经验模态分解(EEMD)降噪与隐马尔科夫模型(HMM)的采煤机摇臂滚动轴承故障诊断方法。采用基于峭度准则的EEMD对采集到的振动信号进行降噪预处理,筛选出包含主要特征频率的本征模态函数(IMF),通过求取IMF信息熵提取出敏感特征集,结合训练好的HMM分类模型,对滚动轴承故障类型进行诊断识别。实验数据分析表明,所提出的基于EEMD降噪和HMM的故障诊断方法可以准确区分滚动轴承故障类型,对于4种状态轴承的识别率达到90%以上,是一种有效的采煤机摇臂滚动轴承故障诊断方法 。
A fault diagnosis method of rocker bearing of shearer based on collective empirical mode decomposition (EEMD) noise reduction and hidden Markov model (HMM) is proposed. The EEMD based on the kurtosis criterion was used to noise reduction of the collected vibration signals to filter out the intrinsic mode function (IMF) containing the main eigenfrequency, extract the sensitive feature set by IMF information entropy, and combine with training HMM classification model, the diagnosis of rolling bearing fault type identification. The experimental data analysis shows that the proposed fault diagnosis method based on EEMD noise reduction and HMM can accurately distinguish the type of rolling bearing fault, and the recognition rate of the four kinds of state bearings is above 90%, which is an effective failure of bearing rocker arm bearing diagnosis method.