摘要
为了有效识别轴承故障,提出基于自适应噪声完备集合经验模态分解与小波阈值联合的卷积神经网络故障诊断模型。首先,运用自适应噪声完备集合经验模态分解算法将采集到的信号分解成本征模态分量,使用小波阈值法对高频的分量进行去噪处理;其次,将去噪后的分量和未去噪的分量进行重构,得到去噪后的信号;最后,基于支持向量机及卷积神经网络建立轴承故障诊断模型,将去噪后的信号进行分类处理。结果表明:基于支持向量机建立的模型准确率可达到88.2%,基于卷积神经网络建立的模型准确率可达到98.5%以上。
To effectively identify bearing faults,a convolutional neural network(CNN)fault diagnosis model based on Complete Ensemble Empirical Mode Decomposition with Adaptive Noise(CEEMDAN)and Wavelet threshold(WT)is proposed.First,the CEEMDAN algorithm is used to decompose the collected signal into Intrinsic Mode Function(IMF),and the Wavelet Threshold method is used to denoise the high-frequency IMF components.Secondly,the denoised IMF component and the un-denoised IMF component are reconstructed to obtain the denoised signal.Finally,bearing fault diagnosis models are established based on Support Vector Machine(SVM)and Convolutional Neural Network(CNN)respectively.The denoised signals are classified and processed.The results show that the accuracy of the model based on Support Vector Machine(SVM)can reach 88.2%.The accuracy of the model based on Convolutional Neural Network(CNN)can reach more than 98.5%.
作者
齐佳宝
王琳
刘劲涛
李家奇
顾渝林
朱怡波
陈冀驰
QI Jiabao;WANG Lin;LIU Jintao;LI Jiaqi;GU Yulin;ZHU Yibo;CHEN Jichi(School of Energy,Power and Nuclear Technology Engineering,Shenyang Institute of Engineering,Shenyang 110136,Liaoning Province;School of Mechanica Engineering,Shenyang Institute of Engineering,Shenyang 110136,Liaoning Province;School of Mechanical Engineering,Shenyang University of Technology,Shenyang 110870,Liaoning Province)
出处
《沈阳工程学院学报(自然科学版)》
2025年第1期84-90,共7页
Journal of Shenyang Institute of Engineering:Natural Science
基金
国家自然科学基金(NSFC 62001312、62101355)
辽宁省联合基金项目面上资助项目(2023-MSLH-214)
辽宁省教育厅科学研究项目(JYTMS20230315)
辽宁省教育厅科学研究项目(LJKZZ20220139)。
作者简介
齐佳宝(2000-),男,辽宁葫芦岛人,硕士研究生。;通讯作者:王琳(1980-),女,辽宁沈阳人,教授,硕士生导师,博士,主要从事信号处理方面的研究。
