Unsupervised Health Monitoring Methods for Truss Robots in Automobile Assembly Workshops

doi:10.3901/JME.2025.04.044

Abstract

Abstract: Truss robots are an important part of automated automotive assembly lines. Real-time monitoring and fault diagnosis of the running condition of truss robots is an important method of improving production efficiency and ensuring the reliability of production lines. In order to solve the problem of scarcity of fault instances in the actual production process, an unsupervised health monitoring method is proposed. First, running data is divided based on the power curves, then the samples for training dataset is generated based on multi-source information fusion method. Then a feature extraction model based on stacked sparse autoencoder is established for obtaining low-dimensional data that reflects critical information during running process. Finally, a one-class classification model based on support vector data description is trained for anomaly detection. In order to verify the effectiveness of the proposed method, validation experiments are conducted on Zollern truss robots in BMW body shop. The results show that the accuracy of the proposed monitoring method reach 91.33% and 99.26% respectively on monitoring tasks of truss robot X-axis and Z-axis geared motors and the proposed method have obvious performance advantages over other unsupervised monitoring methods.

Key words: health monitoring, truss robot, unsupervised learning, multi-source information fusion

CLC Number:

TP242

LIU Kuo, CUI Yiming, YANG Xu, LI Mingyu, LI Kai, WANG Yongqing. Unsupervised Health Monitoring Methods for Truss Robots in Automobile Assembly Workshops[J]. Journal of Mechanical Engineering, 2025, 61(4): 44-54.

References

[1] 郭志良. 数控车床桁架机器人上下料系统设计研究[D]. 大连:大连理工大学，2019. GUO Zhiliang. Design and study of gantry robot system for loading and unloading of CNC lathe[D]. Dalian :Dalian University of Technology，2019.
[2] GAO Z，CECATI C，DING X S. A survey of fault diagnosis and fault-tolerant techniques -part I:fault diagnosis with model-based and signal-based approaches[J]. IEEE Trans. Industrial Electronics，2015，62(6):3757-3767.
[3] 乔石，刘阔，都书博，等. 基于功率信息的航空发动机叶片铣削刀具监测试验研究[J]. 航空制造技术，2021，64(16):87-92，110. QIAO Shi，LIU Kuo，DU Shubo，et al. Research on monitoring test of milling tool for aero-engine blade based on power information[J]. Aeronautical Manufacturing Technology，2021，64(16):87-92,110.
[4] SHAO H，WANG H，ZHAO X. A cutting power model for tool wear monitoring in milling[J]. International Journal of Machine Tools and Manufacture，2004，44(14):1503-1509.
[5] LI W，PENG M，WANG Q. Fault identification in PCA method during sensor condition monitoring in a nuclear power plant[J]. Annals of Nuclear Energy，2018，121:135-145.
[6] 施耀华，冯延晖，任铭，等. 融合电流和振动信号的永磁同步风电系统变流器故障诊断方法研究[J]. 中国电机工程学报，2020，40(23):7750-7760. SHI Yaohua，FENG Yanhui，REN Ming，et al. Study on fault diagnosis method of converter in permanent magnet synchronous wind power system by fusion of current and vibration Signals[J]. Proceedings of the CSEE，2020，40(23):7750-7760.
[7] 叶壮，余建波. 基于多通道一维卷积神经网络特征学习的齿轮箱故障诊断方法[J]. 振动与冲击，2020，39(20):55-66. YE Zhuang，YU Jianbo. Gearbox fault diagnosis based on feature learning of multi-channel one -dimensional convolutional neural network[J]. Journal of Vibration and Shock，2020，39(20):55-66.
[8] 宫文峰，陈辉，张美玲，等. 基于深度学习的电机轴承微小故障智能诊断方法[J]. 仪器仪表学报，2020，41(1):195-205. GONG Wenfeng，CHEN Hui，ZHANG Meiling，et al. Intelligent diagnosis method for incipient fault of motor bearing based on deep learning[J]. Chinese Journal of Scientific Instrument，2020，41(1):195-205.
[9] 李聪波，王睿，张友，等. 基于迁移学习的离心鼓风机故障预警方法[J]. 中国机械工程，2021，32(17):2090-2099，2107. LI Congbo，WANG Rui，ZHANG You，et al. A novel fault early warning method for centrifugal blowers based on transfer learning[J]. China Mechanical Engineering，2021，32(17):2090-2099，2107.
[10] CHENG Yiwei，LIN Manxi，WU Jun，et al. Intelligent fault diagnosis of rotating machinery based on continuous wavelet transform-local binary convolutional neural network[J]. Knowledge-Based Systems，2021，216:106796.
[11] KHORRAM A，KHALOOEI M，REZGHI M. End-to-end CNN + LSTM deep learning approach for bearing fault diagnosis[J]. Applied Intelligence，2020，51(2):736-751.
[12] ZHANG Chen，HU Di，YANG Tao. Anomaly detection and diagnosis for wind turbines using long short-term memory-based stacked denoising autoencoders and XGBoost[J]. Reliability Engineering and System Safety，2022，222:108445.
[13] SUN Peng，LI Jian，WANG Caisheng，et al. A generalized model for wind turbine anomaly identification based on SCADA data[J]. Applied Energy，2016，168:550-567.
[14] ZHOU Funa，YANG Shuai，FUJITA H，et al. Deep learning fault diagnosis method based on global optimization GAN for unbalanced data[J]. Knowledge-Based Systems，2020，187:104837.
[15] YOSHUA B，AARON C，PASCAL V. Representation learning:a review and new perspectives.[J]. IEEE transactions on pattern analysis and machine intelligence，2013，35(8):1798-1828.
[16] OLSHAUSEN A B，FIELD J D. Sparse coding with an overcomplete basis set:A strategy employed by V1?[J]. Vision Research，1997，37(23):3311-3325.
[17] TAX M D，DUIN P R. Support vector data description[J]. Machine Learning，2004，54(1):45-66.
[18] 牛康廷. 桁架机器人运动学建模、定位误差分析及实验研究[D]. 合肥:合肥工业大学，2016. NIU Kangting. Truss robot’s kinematics modeling，positioning error analysis and experimental study[D]. Hefei:Hefei University of Technology，2016.
[19] 刘飞，刘军，何彦. 基于参考功率曲线的大型工件加工进度状态信息的自动采集新方法[J]. 机械工程学报，2009，45(10):111-117. LIU Fei，LIU Jun，HE Yan. Automatic collection method of machining progress information for large-size workpieces based on reference power curve[J]. Journal of Mechanical Engineering，2009，45(10):111-117.
[20] SCHAFER R W. What is a savitzky-golay filter?[J]. IEEE Signal Processing Magazine，2011，28(4):111-117.