EEMD与HMM在齿轮故障诊断方法中的研究

为了解决EMD方法存在的模态混叠的问题,更加精确有效的利用振动信号进行齿轮的故障识别和诊断,提出一种将总体平均经验模态分解(EEMD)和隐马尔科夫模型(HMM)结合的齿轮故障诊断方法。首先对采集到的原始齿轮振动信号进行EEMD处理,获得包含主要故障信息的各阶固有模态函数(IMF)分量,以能量为元素,提取并构造特征向量,对特征向量进行HMM模型训练和诊断测试,来识别齿轮的工作状态和故障类型,实验结果表明,该方法可以有效提高齿轮的故障诊断准确率和精度。

Research on EEMD and HMM in Gear Fault Diagnosis Method

In order to solve the problem of mode mixing in EMD method,and to use vibration signal for gear fault identification and diagnosis more accurately and effectively,a combination of overall mean Ensemble Empirical mode decomposition(EEMD)and Hidden Markov Model(HMM)is proposed.This is a new gear fault diagnosis method.Firstly,EEMD processing is performed on the collected original gear vibration signal to obtain the various natural mode function(IMF)components including the main fault information.The energy vector is used to extract and construct the feature vector,and the HMM model training and diagnosis is performed on the feature vector.And we use this test on feature vector to identify the working state of the gear and the type of fault.The experimental results show that the method can effectively improve the accuracy and accuracy of the fault diagnosis of the gear.