《计算机应用》唯一官方网站 ›› 2021, Vol. 41 ›› Issue (11): 3192-3199.DOI: 10.11772/j.issn.1001-9081.2021010046

• 人工智能 • 上一篇    下一篇

基于动作周期退化相似性度量的机械轴健康指标构建与剩余寿命预测

周玉彬1, 肖红1, 王涛2, 姜文超1,3(), 熊梦3, 贺忠堂3   

  1. 1.广东工业大学 计算机学院,广州 510006
    2.广东工业大学 自动化学院,广州 510006
    3.中国科学院 云计算产业技术创新与育成中心,广东 东莞 523808
  • 收稿日期:2021-01-11 修回日期:2021-05-14 接受日期:2021-05-18 发布日期:2021-11-29 出版日期:2021-11-10
  • 通讯作者: 姜文超
  • 作者简介:周玉彬(1997—),女,湖南常德人,硕士研究生,主要研究方向:工业设备的故障诊断与预测
    肖红(1972—),女,湖北咸宁人,副 教授,博士,主要研究方向:智能制造与工业大数据、机器人与人工智能
    王涛(1983—),男,湖北荆州人,副教授,博士,主要研究方向:物联网、 大数据
    姜文超(1977—),男,山东潍坊人,副教授,博士,CCF会员,主要研究方向:云计算、大数据、复杂网络、图数据处理
    熊梦(1985—), 男,湖北咸宁人,高级工程师,主要研究方向:云计算、大数据、工业互联网
    贺忠堂(1971—),男,江苏徐州人,高级工程师,博士,主要研究方 向:云计算、大数据、政府云服务平台、城市公共安全云管理。
  • 基金资助:
    国家重点研发计划项目(2018YFB1004202);国家自然科学基金委员会-广东省人民政府联合基金资助项目(U2001201);广东省自然科学基金面上项目(2020A1515010890);广东省科技计划项目(2019B010139001);广州市科技计划项目(201902020016);2019年佛山市核心技术攻关项目(1920001001367)

Health index construction and remaining useful life prediction of mechanical axis based on action cycle degradation similarity measurement

Yubin ZHOU1, Hong XIAO1, Tao WANG2, Wenchao JIANG1,3(), Meng XIONG3, Zhongtang HE3   

  1. 1.School of Computer Science,Guangdong University of technology,Guangzhou Guangdong 510006,China
    2.School of Automation,Guangdong University of technology,Guangzhou Guangdong 510006,China
    3.Cloud Computing Center,Chinese Academy of Sciences,Dongguan Guangdong 523808,China
  • Received:2021-01-11 Revised:2021-05-14 Accepted:2021-05-18 Online:2021-11-29 Published:2021-11-10
  • Contact: Wenchao JIANG
  • About author:ZHOU Yubin, born in 1997, M. S. candidate. Her research interests include fault diagnosis and prediction of industrial equipment
    XIAO Hong,born in 1972,Ph. D.,associate professor. Her research interests include intelligent manufacturing and industrial big data,robot and artificial intelligence
    WANG Tao,born in 1983,Ph. D.,associate professor. His research interests include internet of things,big data
    JIANG Wenchao,born in 1977,Ph. D.,associate professor. His research interests include cloud computing,big data,complex network, graph data processing
    XIONG Meng, born in 1985, senior engineer. His research interests include cloud computing,big data,industrial internet
    HE Zhongtang, born in 1971, Ph. D., senior engineer. His research interests include cloud computing,big data,government
  • Supported by:
    the National Key Research and Development Program of China(2018YFB1004202);the National Natural Science Foundation of China-People’s Government of Guangdong Province Joint Fund(U2001201);the Surface Program of Natural Science Foundation of Guangdong Province(2020A1515010890);the Science and Technology Program of Guangdong Province(2019B010139001);the Science and Technology Program of Guangzhou(201902020016);the Key Technology Project of Foshan in 2019(1920001001367)

摘要:

针对工业机器人机械轴健康管理中检测效率和精准度较低的问题,提出了一种机械轴运行监控大数据背景下的基于动作周期退化相似性度量的健康指标(HI)构建方法,并结合长短时记忆(LSTM)网络进行机器人剩余寿命(RUL)的自动预测。首先,利用MPdist关注机械轴不同动作周期之间子周期序列相似性的特点,并计算正常周期数据与退化周期数据之间的偏离程度,进而构建HI;然后,利用HI集训练LSTM网络模型并建立HI与RUL之间的映射关系;最后,通过MPdist-LSTM混合模型自动计算RUL并适时预警。使用某公司六轴工业机器人进行实验,采集了加速老化数据约1 500万条,对HI单调性、鲁棒性和趋势性以及RUL预测的平均绝对误差(MAE)、均方根误差(RMSE)、决定系数(R2)、误差区间(ER)、早预测(EP)和晚预测(LP)等指标进行了实验测试,将该方法分别与动态时间规整(DTW)、欧氏距离(ED)、时域特征值(TDE)结合LSTM的方法,MPdist结合循环神经网络(RNN)和LSTM等方法进行比较。实验结果表明,相较于其他对比方法,所提方法所构建HI的单调性和趋势性分别至少提高了0.07和0.13,RUL预测准确率更高,ER更小,验证了所提方法的有效性。

关键词: MPdist, 长短时记忆网络, 相似性度量, 健康指标构建, 剩余寿命预测

Abstract:

Aiming at the problems of low detection efficiency and accuracy in the health management process of industrial robot axis, a new Health Index (HI) construction method based on action cycle degradation similarity measurement under the background of mechanical axis operation monitoring big data was proposed, and the robot Remaining Useful Life (RUL) prediction was carried out by combining Long Short-Term Memory (LSTM) network. Firstly, MPdist was used to focus on the similarity features of sub-cycle sequences between different action cycles of mechanical axis, and the deviation distance between normal cycle data and degradation cycle data was calculated, so that the HI was constructed. Then, the LSTM network model was trained by HI set, and the mapping relationship between HI and RUL was established. Finally, the MPdist-LSTM hybrid model was used to automatically calculate the RUL and give early warning in time. The six-axis industrial robot of a company was used to carry the experiments, and about 15 million pieces of data were collected. The monotonicity, robustness and trend of HI and Mean Absolute Error (MAE), Root Mean Square Error (RMSE), R-Square (R2), Error Range (ER), Early Prediction (EP) and Late Prediction (LP) of RUL were tested. The proposed method were compared with the methods such as Dynamic Time Warping (DTW), Euclidean Distance (ED), Time Domain Eigenvalue (TDE) combined with LSTM, MPdist combined with RNN and LSTM. The experimental results show that, compared with other comparison methods, the proposed method has the HI monotonicity and trend higher by at least 0.07 and 0.13 respectively, the higher RUL prediction accuracy, and the smaller ER, which verifies the effectiveness of the proposed method.

Key words: MPdist, Long Short-Term Memory (LSTM) network, similarity measurement, Health Index (HI) construction, Remaining Useful Life (RUL) prediction

中图分类号: