基于可解释时空图卷积网络的多传感数据融合诊断方法

doi:10.3901/JME.2024.12.158

机械工程学报 ›› 2024, Vol. 60 ›› Issue (12): 158-167.doi: 10.3901/JME.2024.12.158

• 特邀专栏：可解释可信AI驱动的智能监测与诊断 • 上一篇下一篇

扫码分享

基于可解释时空图卷积网络的多传感数据融合诊断方法

温楷儒¹, 陈祝云^2,3, 黄如意^1,2, 李东鹏¹, 李巍华^2,3

1. 华南理工大学吴贤铭智能工程学院广州 511442;
2. 人工智能与数字经济广东省试验室广州 510335;
3. 华南理工大学机械与汽车工程学院广州 510641

收稿日期:2023-08-01 修回日期:2024-02-01 出版日期:2024-06-20 发布日期:2024-08-23
作者简介:温楷儒,男,1998年出生,博士研究生。主要研究方向为智能故障诊断,深度学习理论及应用。E-mail:202110190432@scut.edu.cn;李巍华(通信作者),男,1973年出生,博士,教授,博士研究生导师。主要研究方向为工业智能、工业大数据、数字孪生、装备智能运维、预测性维护与健康管理、汽车智能驾驶(环境感知、路径规划与决策)。E-mail:whlee@scut.edu.cn
基金资助:
国家自然科学基金(52205100,52275111,52205101,U23A20620)、广东省基础与应用基础研究基金自然科学基金面上(2023A1515012856)和中国博士后科学基金(2022M711197)资助项目。

Multi-sensor Data Fusion Diagnosis Method Based on Interpretable Spatial-temporal Graph Convolutional Network

WEN Kairu¹, CHEN Zhuyun^2,3, HUANG Ruyi^1,2, LI Dongpeng¹, LI Weihua^2,3

1. Shien-Ming Wu School of Intelligent Engineering, South China University of Technology, Guangzhou 511442;
2. Guangdong Artificial Intelligent and Digital Economy Laboratory (Guangzhou), Guangzhou 510335;
3. School of Mechanical & Automotive Engineering, South China University of Technology, Guangzhou 510641

Received:2023-08-01 Revised:2024-02-01 Online:2024-06-20 Published:2024-08-23

摘要/Abstract

摘要： 随着大数据和人工智能技术的发展，深度学习及其衍化算法在故障诊断领域取得了丰硕成果并得到了广泛应用。然而，智能故障诊断方法也存在着诸多挑战：① 现有的多传感数据融合方法难以充分挖掘多传感器间的时空信息来优化诊断性能；② 智能诊断模型的决策过程可解释性弱，无法满足实际工业场景下的可靠性要求。因此，提出一种基于可解释时空图卷积网络的多传感数据融合诊断方法。首先，通过构建门控卷积层以学习并增强时间特征；其次，结合传感器空间布局关系和图卷积网络多传感信息融合能力，学习并提取多传感器数据的空间特征，并通过行星齿轮箱故障诊断任务验证了模型有效性；最后，利用改进梯度类激活映射算法分析各传感器数据对模型诊断决策过程的重要度，从而提升模型更新过程的可解释性。试验结果表明，所提方法不仅具有较好的诊断性能，而且为多源传感融合诊断提供了一种有效的事后解释方法。

关键词: 智能故障诊断, 图卷积网络, 可解释性人工智能, 行星齿轮箱

Abstract: With the development of big data and artificial intelligence technologies, deep learning and its derived algorithms have achieved fruitful results and have been widely used in the field of fault diagnosis. However, intelligent fault diagnosis methods also face many challenges：1) existing multi-sensor data fusion methods have difficulty fully exploring the spatial-temporal information between multiple sensors to optimize diagnostic performance; 2) the decision-making process of intelligent diagnostic models has weak interpretability and cannot meet the reliability requirements in actual industrial scenarios. Therefore, we propose a multi-sensor data fusion diagnostic method based on an interpretable spatial-temporal graph convolutional network(ISTGCN). Firstly, a gate convolution layer is constructed to learn and enhance time features. Secondly, by combining the spatial layout relationship of sensors and the multi-sensor information fusion ability of graph convolutional networks, the spatial features of multi-sensor data are learned and extracted, and the effectiveness of the model is verified through the planetary gearbox fault diagnosis task. Finally, the improved gradient-based activation mapping algorithm is used to analyze the importance of each sensor data to the model diagnostic decision-making process, thereby improving the interpretability of the model update process. The experimental results show that the proposed method not only has good diagnostic performance but also provides an effective post-explanation method for multi-source sensor fusion diagnosis.

Key words: intelligent fault diagnosis, graph convolutional networks, explainable artificial intelligence, planetary gearboxes

中图分类号:

TG156

温楷儒, 陈祝云, 黄如意, 李东鹏, 李巍华. 基于可解释时空图卷积网络的多传感数据融合诊断方法[J]. 机械工程学报, 2024, 60(12): 158-167.

WEN Kairu, CHEN Zhuyun, HUANG Ruyi, LI Dongpeng, LI Weihua. Multi-sensor Data Fusion Diagnosis Method Based on Interpretable Spatial-temporal Graph Convolutional Network[J]. Journal of Mechanical Engineering, 2024, 60(12): 158-167.

参考文献

[1] 黄如意，李霁蒲，王震，等. 基于多任务学习的装备智能诊断与寿命预测方法[J]. 中国科学：技术科学，2022，52：123-137. HUANG Ruyi，LI Jipu，WANG Zhen，et al. Intelligent diagnostic and prognostic method based on multitask learning for industrial equipment[J]. Journal of Mechanical Engineering，2022，52：123-137.
[2] LI W，HUANG R，LI J，et al. A perspective survey on deep transfer learning for fault diagnosis in industrial scenarios：Theories，applications and challenges[J]. Mechanical Systems and Signal Processing，2022，167：108487.
[3] 陈祝云，钟琪，黄如意，等. 基于增强迁移卷积神经网络的机械智能故障诊断[J]. 机械工程学报，2021，57(21)：96-105. CHEN Zhuyun，ZHONG Qi，HUANG Ruyi，et al. Intelligent fault diagnosis for machinery based on enhanced transfer convolutional neural network[J]. Journal of Mechanical Engineering，2021，57(21)：96-105.
[4] 孙文珺，邵思羽，严如强. 基于稀疏自动编码深度神经网络的感应电动机故障诊断[J]. 机械工程学报，2016，52(9)：65-71. SUN Wenjun，SHAO Siyu，YAN Ruqiang. Induction motor fault diagnosis based on deep neural network of sparse auto-encoder[J]. Journal of Mechanical Engineering，2016，52(9)：65-71.
[5] XIAO Y，SHAO H，HAN S Y，et al. Novel joint transfer network for unsupervised bearing fault diagnosis from simulation domain to experimental domain[J]. IEEE/ASME Transactions on Mechatronics，2022，27(6)：5254-5263.
[6] HE F，YE Q. A bearing fault diagnosis method based on wavelet packet transform and convolutional neural network optimized by simulated annealing algorithm[J]. Sensors，2022，22(4)：1410.
[7] 严如强，钱宇宁，胡世杰，等. 基于小波域平稳子空间分析的风力发电机齿轮箱故障诊断[J]. 机械工程学报，2014，50(11)：9-16. YAN Ruqiang，QIAN Yuning，HU Shijie，et al. Wind turbine gearbox fault diagnosis based on wavelet domain stationary subspace analysis[J]. Journal of Mechanical Engineering，2014，50(11)：9-16.
[8] 雷亚国，何正嘉，林京，等. 行星齿轮箱故障诊断技术的研究进展[J]. 机械工程学报，2011，47(19)：59-67. LEI Yaguo，HE Zhengjia，LIN Jing，et al. Research advances of fault diagnosis technique for planetary gearboxes [J]. Journal of Mechanical Engineering，2011，47(19)：59-67.
[9] JING L，WANG T，ZHAO M，et al. An adaptive multi-sensor data fusion method based on deep convolutional neural networks for fault diagnosis of planetary gearbox[J]. Sensors，2017，17(2)：414.
[10] DUAN Z，WU T，GUO S，et al. Development and trend of condition monitoring and fault diagnosis of multi-sensors information fusion for rolling bearings：a review[J]. The International Journal of Advanced Manufacturing Technology，2018，96：803-819.
[11] LIU Y，YAN X，ZHANG C，et al. An ensemble convolutional neural networks for bearing fault diagnosis using multi-sensor data[J]. Sensors，2019，19(23)：5300.
[12] GUAN Y，MENG Z，SUN D，et al. 2Mnet：multi-sensor and multi-scale model toward accurate fault diagnosis of rolling bearing[J]. Reliability Engineering & System Safety，2021，216：108017.
[13] WANG X，MAO D，LI X. Bearing fault diagnosis based on vibro-acoustic data fusion and 1D-CNN network[J]. Measurement，2021，173：108518.
[14] YANG C，LIU J，ZHOU K，et al. An improved multi-channel graph convolutional network and its applications for rotating machinery diagnosis[J]. Measurement，2022，190：110720.
[15] BASAK S，SENGUPTA S，WEN S J，et al. Spatio-temporal AI inference engine for estimating hard disk reliability[J]. Pervasive and Mobile Computing，2021，70：101283.
[16] YANG B，LEI Y，JIA F，et al. A polynomial kernel induced distance metric to improve deep transfer learning for fault diagnosis of machines[J]. IEEE Transactions on Industrial Electronics，2019，67(11)：9747-9757.
[17] SELVARAJU R R，COGSWELL M，DAS A，et al. Grad-cam：visual explanations from deep networks via gradient-based localization[J]. International Journal of Computer Vision，2019，128(2)：336-359.
[18] POPE P E，KOLOURI S，ROSTAMI M，et al. Explainability methods for graph convolutional neural networks[J]. IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)，2019，10772-10781.
[19] CHEN Z，XU J，PENG T，et al. GCN-CAM：A new graph convolutional network-based fault diagnosis method with its interpretability analysis[C]// SAFEPROCESS，2021：1-6.
[20] KIPF T N，WELLING M. Semi-supervised classification with graph convolutional networks[J]. arXiv preprint arXiv，2016：1609.02907.
[21] YU B，YIN H，ZHU Z. Spatio-temporal graph convolutional networks：A deep learning framework for traffic forecasting[J]. arXiv preprint arXiv，2017：1709.04875.
[22] VELICKOVIC P，CUCURULL G，CASANOVA A，et al. Graph attention networks[J]. arXiv preprint arXiv，2017：1710.10903.

基于可解释时空图卷积网络的多传感数据融合诊断方法

Multi-sensor Data Fusion Diagnosis Method Based on Interpretable Spatial-temporal Graph Convolutional Network

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 13

编辑推荐

Metrics

本文评价

[1]	王诗彬, 王世傲, 陈雪峰, 黄海, 安波涛, 赵志斌, 刘永泉, 李应红. 可解释性智能监测诊断网络构造及航空发动机整机试车与中介轴承诊断应用[J]. 机械工程学报, 2024, 60(12): 90-106.
[2]	林京, 焦金阳. 可解释机械智能诊断技术的研究进展与挑战[J]. 机械工程学报, 2023, 59(20): 215-224.
[3]	王辉, 徐佳文, 严如强. 基于多尺度注意力深度强化学习网络的行星齿轮箱智能诊断方法[J]. 机械工程学报, 2022, 58(11): 133-142.
[4]	姜洪开, 邵海东, 李兴球. 基于深度学习的飞行器智能故障诊断方法[J]. 机械工程学报, 2019, 55(7): 27-34.
[5]	胡茑庆, 陈徽鹏, 程哲, 张伦, 张宇. 基于经验模态分解和深度卷积神经网络的行星齿轮箱故障诊断方法[J]. 机械工程学报, 2019, 55(7): 9-18.
[6]	熊鹏, 汤宝平, 邓蕾, 赵明航. 基于动态加权密集连接卷积网络的变转速行星齿轮箱故障诊断[J]. 机械工程学报, 2019, 55(7): 52-57.
[7]	雷亚国, 贾峰, 孔德同, 林京, 邢赛博. 大数据下机械智能故障诊断的机遇与挑战[J]. 机械工程学报, 2018, 54(5): 94-104.
[8]	黄奕宏, 丁康, 何国林. 行星传动系统振动信号数学模型及特征频率分析[J]. 机械工程学报, 2016, 52(7): 46-53.
[9]	黄文涛, 付强, 窦宏印. 基于自适应优化品质因子的共振稀疏分解方法及其在行星齿轮箱复合故障诊断中的应用^*[J]. 机械工程学报, 2016, 52(15): 44-51.
[10]	陈小旺,冯志鹏,LIANG Ming. 基于迭代广义同步压缩变换的时变工况行星齿轮箱故障诊断[J]. 机械工程学报, 2015, 51(1): 131-137.
[11]	雷亚国;孔德同;李乃鹏;林京. 自适应总体平均经验模式分解及其在行星齿轮箱故障检测中的应用[J]. , 2014, 50(3): 64-70.
[12]	雷亚国;汤伟;孔德同;林京. 基于传动机理分析的行星齿轮箱振动信号仿真及其故障诊断[J]. , 2014, 50(17): 61-68.
[13]	雷亚国;何正嘉;林京;韩冬;孔德同. 行星齿轮箱故障诊断技术的研究进展[J]. , 2011, 47(19): 59-67.