Research on Gear Surface Damage Recognition Based on Small Sample Deep Learning

Wang Xiaopeng; Hua Hongpeng; Lu Changqing; Peng Kun; Zhong Yuan; Wu Biqiong

doi:10.16578/j.issn.1004.2539.2024.04.014

您当前的位置：

首页 >

文章列表页 >

Research on Gear Surface Damage Recognition Based on Small Sample Deep Learning

Test·Analysis | 更新时间：2024-04-19

- Research on Gear Surface Damage Recognition Based on Small Sample Deep Learning
- Journal of Mechanical Transmission Vol. 48, Issue 4, Pages: 103-108(2024)
- 作者机构：
  
  1.三峡大学水电机械设备设计与维护湖北省重点实验室，湖北宜昌 443002
  2.湖北特种设备检验检测研究院鄂州分院，湖北鄂州 436000
  3.中国长江电力股份有限公司，湖北宜昌 443133
- 作者简介：
- 基金信息：
- DOI：10.16578/j.issn.1004.2539.2024.04.014
  CLC：
- Published：15 April 2024，
  
  Received：08 January 2023，
扫描看全文
王骁鹏,华鸿鹏,陆长清等.基于小样本深度学习的齿轮表面损伤识别研究[J].机械传动,2024,48(04):103-108.

Wang Xiaopeng,Hua Hongpeng,Lu Changqing,et al.Research on Gear Surface Damage Recognition Based on Small Sample Deep Learning[J].Journal of Mechanical Transmission,2024,48(04):103-108.
王骁鹏,华鸿鹏,陆长清等.基于小样本深度学习的齿轮表面损伤识别研究[J].机械传动,2024,48(04):103-108. DOI： 10.16578/j.issn.1004.2539.2024.04.014.

Wang Xiaopeng,Hua Hongpeng,Lu Changqing,et al.Research on Gear Surface Damage Recognition Based on Small Sample Deep Learning[J].Journal of Mechanical Transmission,2024,48(04):103-108. DOI： 10.16578/j.issn.1004.2539.2024.04.014.

摘要

齿轮表面损伤是影响齿轮传动的重要因素，提高齿轮表面损伤的识别效率和准确率极为重要。基于Pytorch架构建立齿轮表面损伤的ResNet识别模型，利用数据增强的方式扩大数据集，使用迁移学习方式优化模型训练，并对比了4种ResNet结构。结果表明，将64张原始图像数据增强后得到的由640张图像组成的数据集不足以满足模型训练对大量数据的需要；使用迁移学习能够提高模型训练速度和准确率，满足齿轮表面损伤的识别要求；ResNet-101模型在本框架中是最优结构。研究对齿轮表面损伤的检测具有重要的科学意义和工程价值。

Abstract

Gear surface damage is an important factor affecting gear transmission. It is extremely important to improve the efficiency and accuracy of gear surface damage identification. ResNet recognition model of gear surface damage is established based on Pytorch architecture

dataset is expanded by means of data enhancement

model training is optimized by means of transfer learning

and four ResNet structures are compared. The results show that the dataset composed of 640 images after the enhancement of 64 original image is not enough to meet the needs of model training for a large amount of data; using transfer learning can improve the speed and accuracy of model training

and meet the requirements of gear surface damage identification; the ResNet-101 model is the optimal structure in this framework. This research has important scientific significance and engineering value for the recognition of gear surface damage.

关键词

卷积神经网络齿轮表面损伤深度学习迁移学习

Keywords

Convolutional neural networkGear surface damageDeep learningTransfer learning

references

仇娇慧，贝绍轶，尹明锋，等.基于改进YOLOv5s的齿轮表面缺陷检测［J］.现代制造工程，2022（3）：104-113.

QIU Jiaohui，BEI Shaoyi，YIN Mingfeng，et al.Gear surface defect detection based on improved YOLOv5s［J］.Modern Manufacturing Engineering，2022（3）：104-113.

李向波，李涛，石博文，等.基于齿面缺陷激光修复几何数学模型的齿轮修复［J］.中国表面工程，2020，33（3）：129-136.

LI Xiangbo，LI Tao，SHI Bowen，et al.Gear repairing based on geometric mathematical model of tooth surface defects by laser［J］.China Surface Engineering，2020，33（3）：129-136.

孙琦，刘新厂，张兵，等.直齿轮系齿根裂纹损伤程度检测方法［J］.振动、测试与诊断，2019，39（2）：327-333，444-445.

SUN Qi，LIU Xinchang，ZHANG Bing，et al.Detection model of spur gear tooth root crack damage degree［J］.Journal of Vibration，Measurement & Diagnosis，2019，39（2）：327-333，444-445.

赵鑫，张秀华，溥江.齿面点蚀对齿轮啮合刚度影响分析［J］.机械强度，2020，42（6）：1424-1429.

ZHAO Xin，ZHANG Xiuhua，PU Jiang.Effect of pitting on meshing stiffness of gears［J］.Journal of Mechanical Strength，2020，42（6）：1424-1429.

毛晓敏，张慧华，纪晓磊，等.基于XFEM与BP神经网络的裂纹智能识别［J］.应用力学学报，2022，39（6）：1158-1167.

MAO Xiaomin，ZHANG Huihua，JI Xiaolei，et al.Intelligent crack identification based on XFEM and BP neural network［J］.Chinese Journal of Applied Mechanics，2022，39（6）：1158-1167.

SRIVASTAVA N，HINTON G，KRIZHECSKY A，et al.Dropout：a simple way to prevent neural networks from overfitting［J］.Journal of Machine Learning Research，2014，15（1）：1929-1958.

郑远攀，李广阳，李晔.深度学习在图像识别中的应用研究综述［J］.计算机工程与应用，2019，55（12）：20-36.

ZHENG Yuanpan，LI Guangyang，LI Ye.Survey of application of deep learning in image recognition［J］.Computer Engineering and Applications，2019，55（12）：20-36.

LECUN Y，BOTTOU L，BENGIO Y，et al.Gradient-based learning applied to document recognition［J］.Proceedings of the IEEE，1998，86（11）：2278-2324.

KRIZHEVSKY A，SUTSKEVER I，HINTON G E.ImageNet classification with deep convolutional neural networks［J］.Communications of the ACM，2017，60（6）：84-90.

SIMONYAN K，ZISSERMAN A.Very deep convolutional networks for large-scale image recognition［J］.CoRR，2014：1409.1556.

HE K M，ZHANG X Y，REN S Q，et al.Deep residual learning for image recognition［C］//2016 IEEE Conference on Computer Vision and Pattern Recognition （CVPR）.IEEE，2016：770-778.

寇皓为，苏燕辰，陈国俊，等.基于图像处理的齿轮损伤检测与评价方法研究［J］.煤矿机械，2021，42（5）：44-46.

KOU Haowei，SU Yanchen，CHEN Guojun，et al.Research on gear damage detection and evaluation method based on image processing［J］.Coal Mine Machinery，2021，42（5）：44-46.

MILTENOVIC A，RAKONJAC I，OARCEA A，et al.Detection and monitoring of pitting progression on gear tooth flank using deep learning［J］.Applied Sciences，2022，12（11）：5327.

ALLAM A，MOUSSA M，TARRY C，et al.Detecting teeth defects on automotive gears using deep learning［J］.Sensors，2021，21（24）：8480.

LU T C.CNN convolutional layer optimization based on quantum evolutionary algorithm［J］.Connection Science，2021，33（3）：482-494.

IOFFE S，SZEGEDY C.Batch normalization：accelerating deep network training by reducing internal covariate shift［C］//Proceedings of the 32nd International Conference on International Conference Machine Learning.JMLR.org，2015，37：448-456.

蒋昂波，王维维.ReLU激活函数优化研究［J］.传感器与微系统，2018，37（2）：50-52.

JIANG Angbo，WANG Weiwei.Research on optimization of ReLU activation function［J］.Transducer and Microsystem Technologies，2018，37（2）：50-52.

蒋芸，张海，陈莉，等.基于卷积神经网络的图像数据增强算法［J］.计算机工程与科学，2019，41（11）：2007-2016.

JIANG Yun，ZHANG Hai，CHEN Li，et al.An image data augmentation algorithm based on convolutional neural networks［J］.Computer Engineering & Science，2019，41（11）：2007-2016.

GUO L，DAI Q.Laplacian regularized low-rank sparse representation transfer learning［J］.International Journal of Machine Learning and Cybernetics，2020，12（3）：807-821.

KHAN A H，CAO X W，LI S，et al.BAS-ADAM：an ADAM based approach to improve the performance of beetle antennae search optimizer［J］.IEEE-CAA Journal of Automatica Sinica，2020，7（2）：461-471.

SMITH L.Cyclical learning rates for training neural networks［C］//2017 IEEE Winter Conference on Applications of Computer Vision （WACV）.IEEE，2017：464-472.

LIU Y，VATS D，FLEGAL J M.Batch size selection for variance estimators in MCMC［J］.Methodology and Computing in Applied Probability，2021，24（1）：65-93.

KIM B，HAN M，SHIM H，et al.A performance comparison of convolutional neural network-based image denoising methods：the effect of loss function on low-dose CT images［J］.Medical Physics，2019，46（9）：3906-3923.

Views

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Target Detection and Robotic Arm Grasp Pose Estimation Based on YOLOv5 and Transfer Learning

Gear Fault Diagnosis Method Based on Deep Transfer Learning

A VMD and CNN Combined Fault Diagnosis Method for Rolling Bearings

Gear Fault Diagnosis based on Distribution Adaptation Layer and Soft Label Learning

Research of Fault Feature Extraction and Diagnosis of Planetary Gear Train based on SPS and CNN

Related Author

Zhang Zhendong

Ruan Guanqiang

Li Wanyan

Gong Tingkai

Wang Xiyang

Liu Shihao

Li Jishun

Li Kui

Related Institution

School of Mechanical Engineering, University of Shanghai for Science and Technology

School of Mechanical Engineering, Shanghai Dianji University

School of Mechanical and Electrical Engineering, Shanghai Donghai Vocational & Technical College

School of Navigation, Nanchang Hangkong University

School of Aircraft Engineering, Nanchang Hangkong University

⁰