Scholarlink

To solve the inconsistent data distribution of samples under varying working conditions and trouble in collecting an enormous number of samples with labels. In this study, a domain adversarial neural network transfer learning model that includes a residual attention mechanism is proposed. First, to extract essential characteristics from the data, a convolutional neural network with fused a residual attention block is utilized. Then, the adversarial idea and the correlation alignment loss are imported for domain adaptation. Finally, the effectiveness of the suggested approach is assessed based on a dataset that is available to the public, revealing higher diagnostic performance when compared to several other approaches.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Intelligent Fault Diagnosis of Domain Adversarial Bearings for Fusion Attention Mechanisms Under Varying Working Conditions

  • 会议论文

  • 发表日期：

    2025年

查看全文
  • Jiajun Pan,
  • Ke Zhang,
  • Bin Jiang,
  • Lihao Ye,
  • Jinfa Xu

摘要

To solve the inconsistent data distribution of samples under varying working conditions and trouble in collecting an enormous number of samples with labels. In this study, a domain adversarial neural network transfer learning model that includes a residual attention mechanism is proposed. First, to extract essential characteristics from the data, a convolutional neural network with fused a residual attention block is utilized. Then, the adversarial idea and the correlation alignment loss are imported for domain adaptation. Finally, the effectiveness of the suggested approach is assessed based on a dataset that is available to the public, revealing higher diagnostic performance when compared to several other approaches.