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  

Interpretable Network Construction for Intelligent Monitoring and Diagnosis,and Application in Inter-shaft Bearing Diagnosis While Aero-engine Test

作　　者：王诗彬[1,2] 王世傲 陈雪峰 黄海[3] 安波涛 赵志斌 刘永泉 李应红 WANG Shibin;WANG Shiao;CHEN Xuefeng;HUANG Hai;AN Botao;ZHAO Zhibin;LIU Yongquan;LI Yinghong(National Key Lab of Aerospace Power System and Plasma Technology,Xi'an Jiaotong University,Xi'an 710049;School of Mechanical Engineering,Xi'an Jiaotong University,Xi'an 710049;Shenyang Engine Research Institute,Aero Engine Corporation of China,Shenyang 110015)

机构地区：[1]西安交通大学航空动力系统与等离子体技术全国重点实验室,西安710049 [2]西安交通大学机械工程学院,西安710049 [3]中国航空发动机集团公司沈阳发动机研究所,沈阳110015

出　　处：《机械工程学报》2024年第12期90-106,共17页Journal of Mechanical Engineering

基　　金：国家自然科学基金(92270111,52122504,92060302);国家科技重大专项(J2019-I-0001-0001)资助项目。

摘　　要：航空发动机故障预测与健康管理是提高航空发动机安全性、可靠性以及经济可承受性的关键技术。基于深度学习的人工智能方法在机械故障诊断领域受到广泛关注并开展了深入研究,但现有深度学习“黑箱算法”的现状仍然存在模型可解释性差、理论基础薄弱等问题。针对航空发动机健康管理与智能运维的迫切需求,提出航空发动机可解释性智能监测诊断网络,并在某型涡扇发动机整机长试试验中验证了异常检测与中介轴承故障诊断的有效性。将发动机振动信号先验信息融入稀疏表示模型,对模型的迭代求解算法进行展开得到结构具有可解释性的核心网络;针对航空发动机异常检测与智能诊断任务构造了基于对抗训练框架的可解释性异常检测子网络和基于特征提取框架的可解释性故障诊断子网络。本文提出的基于迭代算法展开的网络构造框架具备明确的理论基础,即网络设计有依据;稀疏表示模型驱动的可视化方法能够检验网络是否学到了与发动机故障相符的有意义的特征,即学习结果可信任。最后,通过某型涡扇发动机整机长试试验积累的超过500小时的试车数据,验证了本文提出的模型驱动的可解释性智能监测诊断网络在航空发动机异常检测与中介轴承故障诊断方面的有效性与可靠性。Engine health management is the key technology to improve the safety,reliability and economic affordability of aero-engine.The intelligent diagnosis method based on neural networks has achieved great success in mechanical fault diagnosis,but the current network lacks the targeted design of aero-engine due to its"black box"'nature,and has not been confirmed in engineering practice.In view of these problems,this paper proposes an interpretable network construction framework for intelligent diagnosis of aero-engine and verifies it in the real engine test data.The prior information of aero-engine vibration signals is integrated into the sparse representation model,and the iterative solution algorithm of the model is unrolled to obtain an interpretable core network architecture.The interpretable sub-network via adversarial training is constructed for detection tasks,and the interpretable deep feature extraction sub-network is constructed for intelligent fault diagnosis tasks.Therefore,the network architecture proposed in this paper has a clear theoretical basis,that is,ad-hoc interpretability.In addition,a visualization method is proposed to check whether the network has learned meaningful features,making it post-hoc interpretable.The characteristics of both ad-hoc and post-hoc interpretability make the network more credible when applied to aero-engine anomaly detection and fault diagnosis.Finally,in the long-term test data analysis of a real aero-engine,the interpretable network construction proposed in this paper provides an effective and credible results for fault diagnosis of inter-shaft bearings.

关 键 词：航空发动机健康管理 可解释性人工智能 算法展开 异常检测 故障诊断 

分 类 号：TH17[机械工程—机械制造及自动化]