基于改进MobileViT的风力机增速齿轮箱智能故障诊断方法

陈向民; 雷瀚霖; 张亢; 李泳辉; 李博; 姚鹏

doi:doi:10.19912/j.0254-0096.tynxb.2024-1759

English Version Journal Login

您当前的位置：

首页 >

文章列表页 >

基于改进MobileViT的风力机增速齿轮箱智能故障诊断方法

更新时间：2026-03-06

- 基于改进MobileViT的风力机增速齿轮箱智能故障诊断方法
- Vol. 47, Issue 2, Pages: 470-477(2026)
- 作者机构：
- 作者简介：
- 基金信息：
- DOI：doi:10.19912/j.0254-0096.tynxb.2024-1759
  CLC： TK83
- Online First：05 March 2026，
  
  Published：05 March 2026
- 稿件说明：
移动端阅览
陈向民, 雷瀚霖, 张亢, et al. 基于改进MobileViT的风力机增速齿轮箱智能故障诊断方法[J]. 2026, 47(2): 470-477.
DOI：

陈向民, 雷瀚霖, 张亢, et al. 基于改进MobileViT的风力机增速齿轮箱智能故障诊断方法[J]. 2026, 47(2): 470-477. DOI： doi:10.19912/j.0254-0096.tynxb.2024-1759.

摘要

针对传统方法在增速齿轮箱运行条件下无法充分提取全局信息且识别准确率偏低的问题

提出一种改进的MobileViT风力机增速齿轮箱智能故障诊断方法。该方法在数据预处理模块中利用格拉姆矩阵将一维数据转换成二维图像

保持了信号对时间的依赖性;其次

通过异构内核卷积（HetConv）提取故障浅层信息和提供空间位置偏置

并采用带有自注意力机制（SA）的视觉转换器（ViT）对故障信息进行全局特征提取;最后

使用全连接层对故障进行识别。采用轴承故障数据集和齿轮故障数据集对所提方法进行验证

结果表明：所提方法对齿轮箱状态的识别准确率均在99%以上

优于其他常用网络;且模型参数量更低

模型更小。

Abstract

The fault diagnosis of speed increasing gearbox is of great significance for ensuring the reliable operation of doubly fed wind turbines. However

traditional method cannot fully extract the global information and the recognition accuracy is low under the operating conditions of speed increasing gearbox. In order to tackle this issue an intelligent fault diagnosis technique for the wind turbine's speed-increasing gearbox is proposed in this paper

based on advanced mobile ViT. In this method

the Gram matrix is used to convert the one-dimensional data into a two-dimensional image in the data preprocessing module

which maintains the dependence of the signal on time. Then

heterogeneous kernel-based convolutions （HetConv） are employed to extract the fault local information and also provide spatial position bias

and then the Vision transformer （ViT） with Self-Attention is utilized to extract the global features of the fault information. Finally

fault identification is executed according to the output of the fully connected layer. The experimental results show that the average accuracy of the proposed method is high than 99%

which is better than other commonly used networks. Moreover

the number of model parameters is lower

and the model is smaller.

关键词

Keywords

references

李录平, 刘忠, 张世海, 等. 动力设备故障诊断方法与案例分析[M]. 北京: 中国电力出版社, 2020.

宋礼睿, 崔权维, 周建星, 等. 风电齿轮箱高速轴轴承疲劳寿命及动态可靠性分析[J]. 太阳能学报, 2023, 44(8): 437-444.

TANG J H, WU J M, HU B B, et al.Towards a fault diagnosis method for rolling bearings with time-frequency region-based convolutional neural network[J]. Machines, 2022, 10(12): 1145.

CHEN Y, ZHENG X X.Fault diagnosis of wind turbine based on multi-signal CNN-GRU model[J]. Proceedings of the Institution of Mechanical Engineers, part A: journal of power and energy, 2023, 237(5): 1113-1124.

ZHAO Y, ZHONG Z D, ZHANG H B, et al.Fault diagnosis of rolling bearing based on one-dimensional residual convolution recurrent neural network[J]. Journal of physics: conference series, 2022, 2400(1): 012058.

WEN L, LI X Y, GAO L, et al.A new convolutional neural network-based data-driven fault diagnosis method[J]. IEEE transactions on industrial electronics, 2017, 65(7): 5990-5998.

ZHANG Y, XING K S, BAI R X, et al.An enhanced convolutional neural network for bearing fault diagnosis based on time-frequency image[J]. Measurement, 2020, 157: 107667.

田野, 陈姚节, 张莉, 等. 频域特征提取联合双流CNN的轴承故障诊断方法[J]. 噪声与振动控制, 2024, 44(4): 167-173, 295.

韩康, 战洪飞, 余军合, 等. 基于空洞卷积和增强型多尺度特征自适应融合的滚动轴承故障诊断[J]. 浙江大学学报(工学版), 2024, 58(6): 1285-1295.

葛超, 杨奇睿, 刘佳伟, 等. 基于空洞卷积神经网络与注意力机制GRU的滚动轴承故障诊断[J]. 中国冶金, 2022, 32(4): 99-105, 131.

DOSOVITSKIY A, BEYER L, KOLESNIKOV A, et al.An image is worth 16x16 words: Transformers for image recognition at scale[C]//9th International Conference on Learning Representations ICLR 2021, Virtual Online, 2021.

齐萌, 王国强, 石念峰, 等. 基于时频图与视觉Transformer的滚动轴承智能故障诊断方法[J]. 轴承, 2024(10): 115-123.

TANG X Y, XU Z B, WANG Z G.A novel fault diagnosis method of rolling bearing based on integrated vision transformer model[J]. Sensors, 2022, 22(10): 3878.

张埠石, 范红. 针对小样本改进的MobileViT算法[J]. 计算机工程与应用, 2024, 60(22): 251-260.

梁浩鹏, 曹洁, 赵小强. 基于GADF和PAM-Resnet的旋转机械小样本故障诊断方法[J]. 控制与决策, 2023, 38(12): 3465-3472.

SHAO S Y, MCALEER S, YAN R Q, et al.Highly accurate machine fault diagnosis using deep transfer learning[J]. IEEE transactions on industrial informatics, 2019, 15(4): 2446-2455.

FU Y, CHEN X, LIU Y, et al.Gearbox fault diagnosis based on multi-sensor and multi-channel decision-level fusion based on SDP[J]. Applied sciences, 2022, 12(15): 7535.

HUANG G, LIU Z, VAN DER MAATEN L, et al. Densely connected convolutional networks[C]//2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR). Honolulu, HI, USA, 2017: 2261-2269.

LIU Z, MAO H Z, WU C Y, et al.A ConvNet for the 2020s[C]//2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). New Orleans, LA, USA, 2022: 11966-11976.

杜康宁, 宁少慧, 邓功也. 基于视觉Transformer的滚动轴承智能故障诊断[J]. 组合机床与自动化加工技术, 2023(4): 96-99.

邓飞跃, 郑守禧, 郝如江. 一种轻量化尺度感知调制Swin Transformer模型的轴箱轴承故障诊断方法[J]. 西安交通大学学报, 2024, 58(9): 83-93.

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

启动阶段风电齿轮箱行星轮径向-双向推力耦合滑动轴承瞬态润滑性能分析

基于振动和转子电流深度融合的风电机组转频估计及故障诊断方法

基于改进MW-KECA算法的风电齿轮箱故障诊断研究

基于双绕组异步电机的并网逆变系统及其多工况控制策略

基于仿生结构的复合材料风电叶片损伤分析

Related Author

李浩

谭建军

朱才朝

费文军

孙章栋

王红霞

王城宇

万书亭

Related Institution

School of Electrical Engineering and Information Engineering, Lanzhou University of Technology

School of Information Engineering, Lanzhou City University

Institute of Electric Power, Inner Mongolia University of Technology

Inner Mongolia Key laboratory of Electrical and Mechanical Control

Inner Mongolia North Long yuan Wind Power Co., Ltd.

AI问答

⁰