基于音频峭度的煤矿旋转机械滚动轴承故障预测方法  

Fault Prediction Method for Rolling Bearings in Coal Mine Rotating Machinery Based on Audio Cliffness

作　　者：汪磊 李敬兆[1,2] 秦晓伟 WANG Lei;LI Jingzhao;QIN Xiaowei(School of Electrical and Information Engineering,Anhui University of Science and Technology,Huainan 232001,China;School of Computer Scinence and Engineering,Anhui University of Science and Technology,Huainan 232001,China)

机构地区：[1]安徽理工大学电气与信息工程学院,安徽淮南232001 [2]安徽理工大学计算机科学与工程学院,安徽淮南232001

出　　处：《煤炭技术》2022年第2期173-176,共4页Coal Technology

基　　金：国家自然科学基金项目(51874010,61170060);北京理工大学高精尖机器人开放性研究项目(2018IRS16);物联网关键技术研究创新团队项目(201950ZX003)。

摘　　要：滚动轴承作为煤矿旋转设备中至关重要的机械元件,对其早期故障进行快速有效的诊断与预测可保证矿山开采的稳定性。针对提升机等旋转设备滚动轴承,采用非接触式测量仪器采集轴承工作时的音频信号,通过预加重,分帧加窗及峭度计算提取声音信号的时域特征,并基于萤火虫算法优化的卷积-长短期记忆(CNN-LSTM)神经网络完成音频特征的输出预测。实验结果表明,设计的神经网络模型可对轴承音频的分帧峭度数据进行较为精确的预测拟合,在设定的峭度安全阈值下,该模型能实现滚动轴承早期故障的准确预知。Rolling bearings are crucial mechanical components in coal mining rotating equipment.A rapid and effective diagnosis and prediction of their early failure can ensure the stability of mining.Aiming at the rolling bearings of rotating equipment such as hoists,the audio signals of the bearings are collected by non-contact measuring instruments.Using pre-emphasis,frame-wise windowing and cliff calculation,the time-domain features of the audio signals are extracted.Then,a convolutional-long short-term memory(CNN-LSTM)neural network is optimised based on the firefly algorithm to complete the output prediction of the audio features.The experimental results show that the designed neural network model can make a more accurate prediction fit to the framed cragging data of bearing audio.Under the set safety threshold of cliffness,the model can achieve the accurate prediction of early failure of rolling bearings.

关 键 词：煤矿旋转机械 分帧峭度 混合神经网络模型 萤火虫算法 故障预测 

分 类 号：TD407[矿业工程—矿山机电]