基于CEEMDAN和深度学习的训练器械滚动轴承故障诊断研究  

Research on fault diagnosis of rolling bearing of training equipment based on CEEMDAN and deep learning

作　　者：周闯[1] ZHOU Chuang(Nodern College of Northwest University,Xi’an 710100,China)

机构地区：[1]西北大学现代学院,西安710100

出　　处：《自动化与仪器仪表》2024年第9期101-105,共5页Automation & Instrumentation

摘　　要：针对目前健美操力量训练器械故障诊断方法受噪声干扰以及数据复杂性影响导致诊断准确率低的问题,研究提出了一种结合卷积神经网络和长短时记忆网络的深度学习模型。研究首先利用自适应噪声完备集合经验模态分解对信号进行预处理,然后通过分解振动信号提取关键特征。之后研究利用基于卷积神经网络和长短时记忆网络的深度学习方法对提取的特征进行训练和学习,实现故障类型的自动识别和分类。实验结果表明,研究设计的故障诊断方法对滚动轴承故障的诊断准确率为98.63%,在对不同故障不同等级的情况下,诊断准确率皆达到了90%以上。研究设计方法可实现高精度的故障诊断,提升健美操力量训练器械等设备轴承稳定性和使用周期。A deep learning model combining convolutional neural networks and long short-term memory networks is proposed to address the problem of low diagnostic accuracy caused by noise interference and data complexity in current methods for diagnosing faults in aerobics strength training equipment.The study first preprocesses the signal using adaptive noise complete set empirical mode decomposition,and then extracts key features by decomposing the vibration signal.Afterwards,research will use deep learning methods based on convolutional neural networks and long short-term memory networks to train and learn the extracted features,achieving automatic recognition and classification of fault types.The experimental results show that the fault diagnosis method designed in the study has a diagnostic accuracy of 98.63%for rolling bearing shaft faults,and the diagnostic accuracy has reached over 90%for different levels of faults.Research and design methods can achieve high-precision fault diagnosis,improve the stability and service life of bearings in equipment such as aerobics strength training equipment.

关 键 词：CEEMDAN 深度学习 训练器 滚动轴承 故障 

分 类 号：TP133[自动化与计算机技术—控制理论与控制工程]