• CN:11-2187/TH
  • ISSN:0577-6686

机械工程学报 ›› 2020, Vol. 56 ›› Issue (1): 69-81.doi: 10.3901/JME.2020.01.069

• 机械动力学 • 上一篇    下一篇

基于拉普拉斯特征映射和深度置信网络的半监督故障识别

张鑫1,2, 郭顺生1,2, 李益兵1,2, 江丽1,2   

  1. 1. 武汉理工大学机电工程学院 武汉 430070;
    2. 武汉理工大学湖北省数字制造重点实验室 武汉 430070
  • 收稿日期:2019-02-10 修回日期:2019-07-17 出版日期:2020-01-05 发布日期:2020-03-09
  • 通讯作者: 江丽(通信作者),女,1980年生,博士,讲师,硕士研究生导师。主要研究方向为机械设备状态监测与故障诊断、模式识别。E-mail:happyjl0929@163.com
  • 作者简介:张鑫,男,1994年出生。主要研究方向为故障诊断。E-mail:gzzdzhangxin@163.com
  • 基金资助:
    湖北省自然科学基金(2019CFB565)、中央高校基本科研业务费专项资金(2018IVA022)、国家自然科学基金(51705386,51705385)和湖北省科技支撑计划(2015BAA063、2014BAA032)资助项目。

Semi-supervised Fault Identification Based on Laplacian Eigenmap and Deep Belief Networks

ZHANG Xin1,2, GUO Shunsheng1,2, LI Yibing1,2, JIANG Li1,2   

  1. 1. School of Mechanical and Electronic Engineering, Wuhan University of Technology, Wuhan 430070;
    2. Hubei Digital Manufacturing Key Laboratory, Wuhan University of Technology, Wuhan 430070
  • Received:2019-02-10 Revised:2019-07-17 Online:2020-01-05 Published:2020-03-09

摘要: 针对机械设备故障诊断过程中有标签样本不足,结合流形学习与深度学习的思想,提出了基于拉普拉斯特征映射(Laplacian eigenmap,LE)和深度置信网络(Deep belief network,DBN)的半监督故障识别模型。该模型运用LE算法直接对原始高维振动信号进行特征提取,将低维流形特征输入DBN,利用少量昂贵的有标签样本和大量廉价的无标签样本,二次挖掘故障特征,并构建Soft-max分类器最终识别出机械设备的故障模式。将该半监督模型应用于轴承故障和齿轮裂纹的识别中,试验结果表明,LE算法有效降低了模型的时间复杂度,增强了特征提取的智能性,提高了诊断效率;DBN网络可以充分挖掘故障特性,得到更好的特征表示,提高了分类精度。此外,该模型在不平衡的训练标签下也实现了很好的诊断效果,且适用于多传感器特征融合的诊断,具备实际应用的价值。

关键词: 拉普拉斯特征映射, 深度置信网络, 半监督, 故障诊断

Abstract: Aiming at solving the problem of insufficient labeled samples in mechanical equipment fault diagnosis, a semi-supervised fault recognition model based on Laplace Eigenmap (LE) and Deep Belief Network (DBN) is presented by combining the idea of manifold learning and deep learning. The model utilizes LE algorithm to directly extract the features from the raw high-dimensional vibration signal, and inputs the low-dimensional manifold features into DBN. By using a few expensive labeled samples and lots of cheap unlabeled samples, it excavates fault features for a second time, and finally constructs the Soft-max classification to identify the fault mode of the mechanical equipment. The semi-supervised model is applied to the identification of bearing faults and gear cracks. The test results show that LE algorithm can effectively reduce the time complexity of the model, enhance the intelligence of feature extraction and improve the diagnostic efficiency. DBN network can fully mine fault characteristics, get better feature representation, and improve diagnostic accuracy. In addition, the model also achieves a good diagnosis effect with unbalanced training label, and is applicable to the diagnosis of multi-sensor feature fusion, which has practical application value.

Key words: Laplacian eigenmap, deep belief networks, semi-supervise, fault diagnosis

中图分类号: