摘要
随着现代大型机械设备集成化和自动化程度的不断提高,设备故障诊断领域中存在的数据量大、数据类型多和特征提取困难等问题愈加凸显,对传统的诊断方法提出了挑战。深度置信网络(Deep Belief Network,DBN)凭借其在特征提取与模式识别方面的优势,成为大数据背景下机械设备故障诊断领域中的利器。在阐述DBN实现机械设备故障诊断原理的基础上,分析和总结了该模型在滚动轴承、齿轮箱和旋转机械系统故障诊断中的研究进展,最后对DBN在实现机械设备故障诊断方面面临的挑战和未来值得深入的研究方向进行了探讨与展望。
With the continuous improvement of the integration and automation of modern large-scale mechanical equipment,problems such as large data volume,many data types and difficult feature extraction in equipment fault diagnosis have become more prominent,which poses a challenge to traditional diagnostic methods.The unique advantages of Deep Belief Network(DBN)in feature extraction and pattern recognition make it a better tool for fault diagnosis of mechanical equipment under the background of big data.The principle of DBN to realize the fault diagnosis was described,and the research progress in fault diagnosis of rolling bearing,gearbox and rotating machinery systems was analyzed.Finally,the challenges and future research directions based on DBN in current field were discussed.
作者
李少波
李传江
胡建军
张安思
杨静
Li Shaobo;Li Chuanjiang;Hu Jianjun;Zhang Ansi;Yang Jing(School of Mechanical Engineering,Guizhou University,Guiyang 550025,China;University of South Carolina,Columbia 29208,USA)
出处
《现代制造工程》
CSCD
北大核心
2020年第10期156-162,142,共8页
Modern Manufacturing Engineering
基金
国家自然科学基金资助项目(71962004)
国家智能制造新模式应用项目(工信厅装函[2017]468号)
贵州省科技计划项目(黔科合平台人才[2017]5788,黔科合成果[2018]4706)。
关键词
机械设备
深度置信网络
故障诊断
特征提取
mechanical equipment
deep belief network
fault diagnosis
feature extraction
