基于域泛化D3QN的跨工况故障诊断方法

doi:10.3901/JME.2024.22.165

机械工程学报 ›› 2024, Vol. 60 ›› Issue (22): 165-178.doi: 10.3901/JME.2024.22.165

扫码分享

基于域泛化D3QN的跨工况故障诊断方法

柏林, 何牧耕, 陈兵奎, 刘小峰

重庆大学机械与运载工程学院重庆 400044

收稿日期:2024-01-09 修回日期:2024-06-22 出版日期:2024-11-20 发布日期:2025-01-02
作者简介:柏林,男,1972年出生,博士,教授。主要研究方向为机械设备状态监测与故障诊断、智能测试与仪器。E-mail:bolin0001@aliyun.com;何牧耕,男,1997年出生,硕士。主要研究方向为机械设备状态检测与故障诊断。E-mail:414398805@qq.com;陈兵奎,男,1966年出生,博士,教授,博士研究生导师。主要研究方向为精密传动,不确定性分析及优化设计。E-mail:bkchen@cqu.edu.cn;刘小峰(通信作者),女,1980年出生,博士,教授,博士研究生导师。主要研究方向为设备监测与故障诊断、结构损伤检测。E-mail:liuxfeng0080@126.com
基金资助:
国家科技重大专项(J2019-IV-0001-0068)和国家自然科学基金(52175077，51975067)资助项目。

Domain Generalization D3QN for Machinery Fault Diagnosis Across Different Working Conditions

BO Lin, HE Mugeng, CHEN Bingkui, LIU Xiaofeng

College of Mechanical and Transportation Engineering, Chongqing University, Chongqing 400044

Received:2024-01-09 Revised:2024-06-22 Online:2024-11-20 Published:2025-01-02
About author:10.3901/JME.2024.22.165

摘要/Abstract

摘要： 针对深度强化学习对交互环境的依赖性导致的其在跨工况设备故障诊断中可移植性差的问题，提出一种D3QN (Dueling double deep Q network, D3QN) 域泛化的故障诊断方法。采用自适应权值的最大相关最小冗余特征筛选方法进行特征优化选择，实现数据环境去冗余精化处理；在竞争网络和双Q网络基础上引入了域识别网络，实现工况环境掩蔽下的故障状态信息分离提取；构建基于故障模式类间距的量化奖励矩阵，并结合域辨识奖励设置分治奖励策略，增强智能体对跨工况混叠故障模式的辨识决策能力。齿轮箱故障与轴承故障的跨工况诊断结果表明，能够较好地解决深度强化学习网络对交互环境的依赖性和其在跨工况故障诊断中与环境独立性之间的矛盾问题，实现深度强化模型在不同工况环境中的复用移植，提高深度强化学习在跨域故障诊断中的适用性。

关键词: 故障诊断, 域泛化, 特征筛选, 分治奖励, 深度强化学习

Abstract: To address the problem of poor portability of deep reinforcement learning model in cross-condition fault diagnosis due to its dependence on the interaction environment, a domain generalization D3QN (Domain generalization dueling double deep Q network, DGD3QN)model is proposed for the machinery fault diagnosis across different working conditions. To realize the de-redundancy and refinement of data environment, the adaptive weighted max-relevance-min-redundancy method is utilized to optimize feature selection. The domain recognition network branch is introduced into D3QN network to separate and extract the fault state information from multi-conditions. To enhance the agent’s ability of identifying the overlapping failure modes in the multi-condition, the graded reward strategy is set by combining the domain recognition reward and the quantitative reward matrix constructed based on the inter-class distance of multi-condition failure modes. The experimental results of cross-condition diagnosis of gearbox fault and bearing fault showed that the proposed DGD3QN can better solve the contradiction between the environment dependence of DQN and the independence of cross-condition fault diagnosis on environmental conditions, realize the multiplexing and transplantation of D3QN models in different operating environments and enhance the applicability of DQN in the cross-domain fault diagnosis accuracy.

Key words: fault diagnosis, domain generalization, feature screening, graded reward strategy, deep reinforcement learning

中图分类号:

TH133

柏林, 何牧耕, 陈兵奎, 刘小峰. 基于域泛化D3QN的跨工况故障诊断方法[J]. 机械工程学报, 2024, 60(22): 165-178.

BO Lin, HE Mugeng, CHEN Bingkui, LIU Xiaofeng. Domain Generalization D3QN for Machinery Fault Diagnosis Across Different Working Conditions[J]. Journal of Mechanical Engineering, 2024, 60(22): 165-178.

参考文献

[1] 宋秉睿. 基于机器学习的配电网故障定位及负荷转供研究[D]. 北京：北京交通大学，2021. SONG Bingrui. Research on fault location and load transfer in distribution network based on machine learning[D]. Beijing：Beijing Jiaotong University，2021.
[2] SHANTIA A，TIMMERS R，CHONG Y，et al. Two-stage visual navigation by deep neural networks and multi-goal reinforcement learning[J]. Robotics and Autonomous Systems，2021，138：103731.
[3] LAI Y H，WU T C，LAI C F，et al. Cognitive optimal-setting control of a IoT industrial applications with deep reinforcement learning[J]. IEEE Transactions on Industrial Informatics，2020，17(3)：2116-2123.
[4] IBRAHIM A M，YAU K L A，CHONG Y W，et al. Applications of multi-agent deep reinforcement learning: models and algorithms[J]. Applied Sciences，2021，11(22)：10870.
[5] ZHOU S K，LE H N，LUU K，et al. Deep reinforcement learning in medical imaging：A literature review[J]. Medical Image Analysis，2021，73：102193.
[6] LEE J，KOH H，CHOE H J. Learning to trade in financial time series using high-frequency through wavelet transformation and deep reinforcement learning[J]. Applied Intelligence，2021，51(8)：6202-6223.
[7] DAI W，MO Z，LUO C，et al. Fault diagnosis of rotating machinery based on deep reinforcement learning and reciprocal of smoothness index[J]. IEEE Sensors Journal，2020，20(15)：8307-8315.
[8] WEN L，LI X，GAO L. A new reinforcement learning based learning rate scheduler for convolutional neural network in fault classification[J]. IEEE Transactions on Industrial Electronics，2020，68(12)：12890-12900.
[9] ZHANG W，FENG W，ZHAO H，et al. Rapidly learning bayesian networks for complex system diagnosis：A reinforcement learning directed greedy search approach[J]. IEEE Access，2019，8：2813-2823.
[10] 辛阔，王建国，张文兴. 基于深度Q学习策略的旋转机械故障诊断方法研究[J]. 机电工程，2021，38(10)：1261-1268. XIN Kuo，WANG Jianguo，ZHANG Wenxing. Research on fault diagnosis method of rotating machinery based on deep Q-learning strategy[J]. Mechatronics Engineering，2021，38(10)：1261-1268.
[11] DING Y，MA L，MA J，et al. Intelligent fault diagnosis for rotating machinery using deep Q-network based health state classification：A deep reinforcement learning approach[J]. Advanced Engineering Informatics，2019，42：100977.
[12] 陈仁祥，周君，胡小林，等. 多Agent深度Q学习和模糊积分的行星齿轮箱故障诊断[J]. 振动与冲击，2021，40(11)：147-153. CHEN Renxiang，ZHOU Jun，HU Xiaolin，et al. Multi-agent deep Q-learning and fuzzy integration for planetary gearbox fault diagnosis[J]. Vibration and Shock，2021，40(11)：147-153.
[13] WU Z，JIANG H，LIU S，et al. A deep reinforcement transfer convolutional neural network for rolling bearing fault diagnosis[J]. ISA Transactions，2022，129：505-524.
[14] LI G，WU J，DENG C，et al. Deep reinforcement learning-based online Domain adaptation method for fault diagnosis of rotating machinery[J]. IEEE/ASME Transactions on Mechatronics，2022，27(5)：2796-2805.
[15] FAN S，ZHANG X，SONG Z. Imbalanced sample selection with deep reinforcement learning for fault diagnosis[J]. IEEE Transactions on Industrial Informatics，2021，18(4)：2518-2527.
[16] 康守强，刘哲，王玉静，等. 基于改进DQN网络的滚动轴承故障诊断方法[J]. 仪器仪表学报，2021，42(3)：201-212. KANG Shouqiang，LIU Zhe，WANG Yujing，et al. A rolling bearing fault diagnosis method based on improved DQN network[J]. Journal of Instrumentation，2021，42(3)：201-212.
[17] MNIH V，KAVUKCUOGLU K，SILVER D，et al. Human- level control through deep reinforcement learning[J]. Nature 2015，518(2015)：529-533.
[18] YU Dinga，LIANG Maa，JIAN Ma，et al. Intelligent fault diagnosis for rotating machinery using deep Q-network based health state classification：A deep reinforcement learning approach[J]. Advanced Engineering Informatics 2019，42(2019)：100977
[19] 袁泉，曾文驱，李子涵，等. 基于改进型D3QN深度强化学习的铁路智能选线方法[J]. 铁道科学与工程学报，2022，19(2)：344-350. YUAN Quan，ZENG Wenqu，LI Zihan，et al. Intelligent railroad routing method based on improved D3QN deep reinforceent learning[J]. Journal of Railway Science and Engineering，2022，19(2)：344-350.
[20] 李长松，刘凯，肖先勇，等. 基于条件互信息特征选择法和Adaboost算法的电能质量复合扰动分类[J]. 高电压技术，2019，45(2)：579-585. LI Changsong，LIU Kai，XIAO Xianyong，et al. Power quality composite disturbance classification based on conditional mutual information feature selection method and Adaboost algorithm[J]. High Voltage Technology，2019，45(2)：579-585.
[21] 谢彦红，孙呈敖，李元. 加权动态SVDD在非线性过程监测中的应用研究[J]. 系统仿真学报，2017，29(7)：1506-1513. XIE Yanhong，SUN Chengao，LI Yuan. Research on the application of weighted dynamic SVDD in nonlinear process monitoring[J]. Journal of System Simulation，2017，29(7)：1506-1513.
[22] TANG Z，BO L，LIU X，et al. A semisupervised transferable lstm with feature evaluation for fault diagnosis of rotating machinery[J]. Applied Intelligence，2022，52(2)：1703-1717.
[23] GANIN Y，USTINOVA E，AJAKAN H，et al. Domain adversarial training of neural networks[J]. The Journal of Machine Learning Research，2016，17(1)：2096-2130.

基于域泛化D3QN的跨工况故障诊断方法

Domain Generalization D3QN for Machinery Fault Diagnosis Across Different Working Conditions

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics

本文评价

[1]	王无印, 黄子钊, 庄子龙, 方怀瑾, 秦威. 基于深度强化学习的自动化码头堆场场桥调度方法[J]. 机械工程学报, 2024, 60(6): 44-57.
[2]	狄子钧, 袁东风, 李东阳, 梁道君, 周晓天, 信苗苗, 曹凤, 雷腾飞. 基于多尺度-高效通道注意力网络的刀具故障诊断方法[J]. 机械工程学报, 2024, 60(6): 82-90.
[3]	严如强, 许文纲, 王志颖, 朱启翔, 周峥, 赵志斌, 孙闯, 王诗彬, 陈雪峰, 张军辉, 徐兵. 航空发动机燃油控制系统故障诊断技术研究进展与挑战[J]. 机械工程学报, 2024, 60(4): 3-31.
[4]	张军辉, 刘施镐, 徐兵, 黄伟迪, 吕飞, 黄晓琛. 轴向柱塞泵智能化关键技术研究进展及发展趋势[J]. 机械工程学报, 2024, 60(4): 32-49.
[5]	丁孺琦, 熊文杰, 程敏, 徐兵. 智能化阀口独立电液控制系统安全性能评估[J]. 机械工程学报, 2024, 60(4): 101-112.
[6]	王志颖, 李天福, 许文纲, 孙闯, 张军辉, 徐兵, 严如强. 降噪混合注意力变分自编码器及其在轴向柱塞泵故障诊断中的应用[J]. 机械工程学报, 2024, 60(4): 167-177.
[7]	邵海东, 林健, 闵志闪, 明宇航. 分布外样本干扰下基于改进半监督原型网络的齿轮箱跨域故障诊断[J]. 机械工程学报, 2024, 60(4): 212-221.
[8]	潘海洋, 徐海锋, 郑近德, 童靳于, 张飞斌. 基于双加权不平衡矩阵分类器的机械故障诊断方法[J]. 机械工程学报, 2024, 60(3): 170-180.
[9]	郭景华, 李文昌, 王班, 王靖瑶. 基于深度强化学习的网联混合动力汽车队列控制[J]. 机械工程学报, 2024, 60(2): 262-271.
[10]	赵轲, 叶敏, 王瑞欣, 陆海, 刘孟孟, 邵海东. 基于模糊域自适应的源自由域旋转机械故障诊断方法[J]. 机械工程学报, 2024, 60(18): 43-52.
[11]	李志鹏, 马天雨, 刘金平, 向青松, 唐俊杰. 基于非对称性对抗训练的多源域自适应智能故障诊断方法[J]. 机械工程学报, 2024, 60(18): 76-88.
[12]	汪煜坤, 易彩, 汪浩, 周秋阳, 冉乐, 王靖元. PSD引导的自适应频带划分方法及其在轴承故障诊断中的应用[J]. 机械工程学报, 2024, 60(17): 179-193.
[13]	刘一龙, 李心远, 陈银萍, 成玮, 陈雪峰. 基于峭度曲面极大值的电机轴承保持架故障诊断方法[J]. 机械工程学报, 2024, 60(15): 89-99.
[14]	于小洛, 杨阳, 杜明刚, 何清波, 彭志科. 机械传动系统含偏差频率分量协同检测与分解方法[J]. 机械工程学报, 2024, 60(15): 100-112.
[15]	黄包裕, 张永祥. 一种基于三步法的轴承故障诊断方法[J]. 机械工程学报, 2024, 60(14): 51-68.