| 融合视觉编码与多尺度注意力的轴承故障诊断方法 |
投稿时间：2025-12-28 修订日期：2026-01-15  点此下载全文 |
| 引用本文： |
| 摘要点击次数: 28 |
| 全文下载次数: 0 |
|
| 基金项目:山东省重大科技创新工程项目（2021CXGC011204）；山东省重点研发计划（科技型中小企业创新能力提升工程）项目（2024TSGC0106） |
|
| 中文摘要:滚动轴承是旋转机械核心部件，其运行状态关乎设备安全可靠。复杂工况下振动信号故障特征呈非线性、多尺度特性，传统方法存在特征建模不足、细粒度判别难、微弱故障易被噪声掩盖等问题。为此，本研究提出融合感受野增强模块（RFB）与可变形注意力机制的 RFB-S-DT 轴承故障诊断方法：先将一维振动信号映射为二维图像；再在网络浅层设 RFB 模块实现多尺度感受野自适应增强；最后结合 Swin-Deformable Transformer 构建深层特征提取网络。基于 CWRU 数据集的实验表明，该方法 10 类故障诊断准确率 98.3%、F1 分数 98.1%，优于主流模型且复杂度合理，为复杂工况下轴承细粒度故障诊断提供有效方案。 |
| 中文关键词:滚动轴承故障诊断 GASF 感受野增强 注意力机制 Swin-Deformable Transformer |
| |
| A Rolling Bearing Fault Diagnosis Method Integrating Visual Encoding and Multi-Scale Attention |
|
|
| Abstract:Rolling bearings are core components of rotating machinery, and their operational status is critical to the safety and reliability of equipment. Under complex working conditions, the fault characteristics of vibration signals exhibit nonlinear and multi-scale properties. Traditional methods suffer from insufficient multi-scale feature modeling, difficulty in fine-grained fault discrimination, and weak faults being easily masked by noise. To address these issues, this study proposes an RFB-S-DT bearing fault diagnosis method that integrates the Receptive Field Block (RFB) and deformable attention mechanism. The method first converts one-dimensional vibration signals into two-dimensional images, then embeds the RFB module in the shallow layer of the network to achieve adaptive enhancement of multi-scale receptive fields, and finally constructs a deep feature extraction network combined with Swin-Deformable Transformer. Experimental results based on the CWRU dataset show that the proposed method achieves an accuracy of 98.3% and an F1-score of 98.1% in 10-class fault diagnosis tasks, outperforming mainstream models with reasonable complexity. It provides an effective solution for fine-grained fault diagnosis of bearings under complex working conditions. |
| keywords:rolling bearing fault diagnosis GASF receptive field enhancement attention mechanism Swin-Deformable Transformer |
| 查看全文 查看/发表评论 下载pdf阅读器 |
次