摘要
在现代工业生产中,机械设备智能化水平提高的同时,其结构和功能也日趋复杂。企业在生产效率提高的同时,也对生产安全提出了更高的要求,而传统的旋转传动部件故障诊断方法显然不足以满足要求。因此,文中主要以结合卷积神经网络的旋转传动部件故障智能诊断为研究对象展开了论述,首先介绍了目前机械领域故障诊断发展的现状,以及以卷积神经网络为代表的智能诊断方法所存在的优势,随后对卷积神经网络的基本概念进行了介绍,然后将卷积神经网络在滚动轴承、齿轮箱和转子系统上故障诊断的应用情况进行了总结分类,最后,总结展望了当前旋转传动部件故障诊断领域研究中存在的问题及可能的发展趋势。
In modern industrial production,while the intelligent level of mechanical equipment is improving,its structure and function become more and more complex.While ensuring higher production efficiency,enterprises put forward more requirements for production safety;the traditional fault-diagnosis method for rotating transmission components is obviously not enough to meet the requirements.Therefore,in this article,efforts are made to focus on the intelligent fault diagnosis of rotating transmission components,combined with the convolutional neural network.Firstly,the current development of fault diagnosis in the mechanical field is introduced,as well as the advantages of the intelligent method of fault diagnosis represented by the convolutional neural network.Then,the basic concept of the convolutional neural network is explored.The application of the convolutional neural network in fault diagnosis of rolling bearings,gearboxes and rotor systems is summarized and classified.Finally,efforts are made to review the current problems and future prospects of fault diagnosis of rotating transmission components.
作者
刘磊
李舜酩
陆建涛
LIU Lei;LI Shun-ming;LU Jian-tao(College of Energy and Power Engineering,Nanjing University of Aeronautics and Astronautics,Nanjing 210016)
出处
《机械设计》
CSCD
北大核心
2022年第10期1-8,共8页
Journal of Machine Design
基金
国家科技重大专项项目(2017-IV-0008-0045)——两机专项项目。
关键词
卷积神经网络
深度学习
故障诊断
convolutional neural network
deep learning
fault diagnosis
