面向旋转机械迁移诊断的分层并行网络模型自动创建方法  

Automatic Model Creation Method of Hierarchical Parallel Network Model for Transfer Diagnosis of Rotating Machinery

作　　者：周健 郑联语[1,2] 王艺玮[1,2] 王移川 ZHOU Jian;ZHENG Lianyu;WANG Yiwei;WANG Yichuan(School of Mechanical Engineering and Automation,Beihang University,Beijing 100191;MIIT Key Laboratory of Intelligent Manufacturing Technology for Aeronautics Advanced Equipments,Ministry of Industry and Information Technology,Beijing 100191;Beijing Institute of Precision Mechatronics Control Equipment,Beijing 100076;Aerospace Servo Drive and Transmission Technology Laboratory,Beijing 100076)

机构地区：[1]北京航空航天大学机械工程及自动化学院,北京100191 [2]航空高端装备智能制造技术工业和信息化部重点实验室,北京100191 [3]北京精密机电控制设备研究所,北京100076 [4]航天伺服驱动与传动技术实验室,北京100076

出　　处：《机械工程学报》2022年第22期115-128,共14页Journal of Mechanical Engineering

基　　金：国家重点研发计划(2020YFB1708400);航天伺服驱动与传动技术实验室开发基金(LASAT-2021-01);国家自然科学基金(51805262)资助项目。

摘　　要：针对当前基于深度学习的旋转机械故障诊断方法存在的依赖人工建模经验、需手动调参、试错迭代、面对不同诊断任务需重新创建诊断模型、异构迁移性差等问题,文中提出一种面向旋转机械迁移诊断的分层并行网络模型自动创建方法,可根据不同诊断任务快速自动地搜索出具有异构迁移性能的高精度诊断模型。基于神经结构搜索(Neural architecture search, NAS)与模块化设计的思想,设计了两类包含多层网络并行结构的基础块,区别于逐网络层搜索的模式,以基础块为单位进行搜索提高效率,控制器输出决策序列确定基础块的内部结构,并将其堆叠形成分层并行结构的子模型,根据子模型在诊断任务上的验证结果利用策略梯度算法优化控制器,循环迭代上述过程不断提高子模型的诊断精度。子模型的分层并行结构支撑了其良好的异构迁移性能,此外为解决NAS搜索耗时的瓶颈问题,在子模型训练过程中设置了权值共享机制以提高自动建模效率。所提方法面向四个不同旋转机械故障数据集进行自动建模和异构迁移诊断试验,结果表明针对四个不同诊断任务,所提方法均能高效创建出100%精度的诊断模型,消耗时间313 s到1 601 s不等,并且所创建的子模型在仅用10%目标域数据耗费100 s时间进行微调的条件下,即可面对目标诊断任务达到95%以上的迁移诊断精度。In view of the current deep learning-based fault diagnosis methods of rotating machinery relying on manual modeling experience, requiring manual parameter adjustment and continuous trial-and-error iteration, re-creation of diagnosis models for different diagnostic tasks, an automatic creation method of hierarchical parallel network model for transfer diagnosis of rotating machinery is proposed, which can quickly and automatically search high precision diagnosis models with heterogeneous transfer performance according to different diagnosis tasks. Based on neural architecture search(NAS) and modular design ideas, two types of foundation blocks of parallel structure containing multiple layers are designed, which is different from the traditional NAS method to search layer by layer, but searches based on the foundation blocks. The controller outputs decision sequence to determine the foundation blocks’ structure and stacks them to form a hierarchical parallel candidate model. Then according to the verification results of the candidate model on the diagnosis task, the controller is optimized using the strategy gradient algorithm, and the diagnosis accuracy of the candidate model is continuously improved by iterating the above process. The hierarchical parallel structure of the candidate model supports its good heterogeneous transfer performance. In addition, in order to solve the time-consuming bottleneck problem of NAS method, the weight sharing mechanism is set in the candidate model training process to improve the efficiency of automatic modeling. The proposed method is used to conduct automatic modeling and heterogeneous transfer diagnosis experiments for four different rotating machinery fault datasets, and the results show that the proposed method can efficiently create 100%accuracy diagnosis models for four different diagnosis tasks, consuming 313 s to 1 601 s, and the candidate model can achieve a transfer diagnosis accuracy of more than 95% for the target diagnosis task with only 10% of the target domain d

关 键 词：旋转机械 分层并行网络 模型自动创建 权值共享 迁移诊断 

分 类 号：TH165[机械工程—机械制造及自动化] TG506[金属学及工艺—金属切削加工及机床]