基于EEMD与CNN-BiLSTM的噪声环境下滚动轴承故障诊断方法  

Fault Diagnosis Method for Rolling Bearings Under Noisy Environments Based on EEMD and CNN-BiLSTM

作　　者：李军星 徐行 贾现召[1] 邱明[1] LI Junxing;XU Hang;JIA Xianzhao;QIU Ming(School of Mechatronics Engineering,Henan University of Science and Technology,Luoyang 471003,China;Henan Provincial Collaborative Innovation Center for High-End Bearings,Luoyang 471003,China)

机构地区：[1]河南科技大学机电工程学院,河南洛阳471003 [2]高端轴承河南省协同创新中心,河南洛阳471003

出　　处：《轴承》2025年第2期85-92,共8页Bearing

基　　金：国家自然科学基金资助项目(52005159);河南省科技攻关项目(222102220061);河南省高校青年骨干教师培养计划资助项目(2021GGJS048);河南省青年托举人才项目(2023HYTP050)。

摘　　要：针对滚动轴承在噪声环境中发生故障时,传统深度神经网络容易出现特征提取不充分,过拟合,泛化能力不足的问题,提出一种集成经验模态分解(EEMD)与卷积神经网络-双向长短时记忆网络(CNN-BiLSTM)的故障诊断方法。在信号预处理阶段使用EEMD将噪声环境下的振动信号分解为一系列固有模态函数,降低噪声的影响;在CNN部分的第1层使用大卷积核与多分支结构获得不同的感受野,在每一个分支中随机丢弃一些数据增强模型的抗干扰能力,从而提取到更具泛化能力的多样化特征信息,后续部分使用残差结构,以免网络较深时发生梯度消失的现象,解决深层次网络退化问题;在BiLSTM部分使用2个并行的分支结构,用于增强模型对时序信息的利用,从而提高模型在不同工况和噪声环境下的准确率。使用凯斯西储大学轴承数据集和西安交通大学轴承数据集对所提方法进行验证,并与其他深度学习方法和传统机器学习方法进行对比,结果表明本文方法在多种工况和噪声环境下均取得了优异的故障诊断性能。A fault diagnosis method based on ensemble empirical mode decomposition(EEMD)and convolutional neural network bi-directional long short-term memory network(CNN-BiLSTM)is proposed to address the problems of inadequate feature extraction,overfitting and insufficient generalization ability of traditional deep neural networks when rolling bearings fail under noisy environments.During signal preprocessing stage,EEMD is used to decompose the vibration signals under noisy environments into a series of intrinsic mode functions,reducing the impact of noise.In the first layer of CNN part,large convolutional kernel and multi-branch structure are used to obtain different receptive fields,and some data are randomly discarded in each branch to enhance the anti-interference ability of the model,so as to extract diversified feature information with more generalization ability.In subsequent parts,residual structures are used to avoid the phenomenon of gradient disappearance when the network is deep,solving the problem of deep network degradation.In BiLSTM part,two parallel branch structures are used to enhance the utilization of temporal information,thereby improving the accuracy of the model under different operating conditions and noise environments.Using bearing datasets from Case Western Reserve University and Xi’an Jiaotong University,the proposed method is verified and compared with other deep learning methods and traditional machine learning methods.The results show that the proposed method achieves excellent fault diagnosis performance under various operating conditions and noise environments.

关 键 词：滚动轴承 故障诊断 集成经验模态分解 卷积神经网络 双向长短时记忆神经网络 

分 类 号：TH133.33[机械工程—机械制造及自动化]