基于小波包与隐马尔可夫的矿井提升机主轴故障诊断

Fault Diagnosis of Mine Hoist Spindle Based on Wavelet Packet and HMM

为解决当前矿井提升机主轴故障数据提取困难且诊断方法存在易受干扰、误差大、准确度低等缺点,设计了基于小波包与隐马尔可夫(HMM)的矿井提升机主轴故障诊断模型。该模型预先把主轴振动信号用小波包分解来获取小波包能量,再把高能量频带CEEMD分解,选取相关系数满足条件的IMF分量完成信号重构,通过重构信号来获得特征参数并构建特征向量,然后对每种故障完成HMM训练,构建HMM故障识别库,并把测试样本送入库中完成测试,从而测试模型的准确度。测试数据表明了基于小波包与HMM的故障诊断模型,准确度高、误差小、抗干扰能力强,比较适用于故障诊断。

In order to solve the current mine hoist spindle fault data extraction difficulties and diagnosis methods are easy to be disturbed,large error,low accuracy shortcomings,a mine hoist spindle fault diagnosis model based on wavelet packet and hidden Markov(HMM)was designed.The model in advance the main shaft vibration signals with wavelet packet decomposition to obtain wavelet packet energy,and the high energy band CEEMD decomposition,correlation coefficient to satisfy conditions selected the IMF component complete signal reconstruction,through reconstructing signal to obtain the characteristic parameters and build characteristic vector,and then complete the HMM training on each type of fault,fault identification library build HMM,And the test samples are sent to the library to complete the test,so as to test the accuracy of the model.The test data show that the fault diagnosis model based on wavelet packet and HMM has high accuracy,small error and strong anti-jamming ability,so it is more suitable for fault diagnosis.