机器学习在涡轮机械中的应用进展  

Application Progress of Machine Learning in Turbomachinery

作　　者：刘浩[1,2,3] 李国庆 张深[1,2,3] 李昂 张燕峰 卢新根[1,2,3] LIU Hao;LI Guoqing;ZHANG Shen;LI Ang;ZHANG Yanfeng;LU Xingen(Key Laboratory of Light-Duty Gas-Turbine/Institute of Engineering Thermophysics,Chinese Academy of Sciences,Beijing 100190,China;School of Aeronautics and Astronautics,University of Chinese Academy of Sciences,Beijing 100049,China;Innovation Academy for Light-duty Gas Turbine,Chinese Academy of Sciences,Beijing 100190,China)

机构地区：[1]中国科学院工程热物理所/轻型动力重点实验室,北京100190 [2]中国科学院大学航空宇航学院,北京100049 [3]轻型动力创新研究院,北京100190

出　　处：《工程热物理学报》2023年第4期938-951,共14页Journal of Engineering Thermophysics

基　　金：国家自然科学基金(No.51976214);航空发动机及燃气轮机基础科学中心项目(No.P2022-B-II-006-001)。

摘　　要：涡轮机械在工业中广泛应用,是实现能源转化利用的重要载体。随着涡轮机械的不断发展,利用数值仿真和实验手段对其性能改进空间有限,难以满足与日俱增的工程需要。近年来,随着人工智能和数据驱动技术的快速发展,机器学习作为其主要分支,备受关注,在诸多领域展现了巨大的应用潜力。将机器学习应用于涡轮机械,对于涡轮机械部件优化设计具有效率高、可泛化等优点,同时使涡轮机械部件寿命预测、健康监测和故障诊断变得更加智能高效,为涡轮机械发展带来了新的契机。综述了机器学习的基本原理以及经典机器学习算法,对机器学习在涡轮机械部件如叶片、轮盘和密封组件的研究进展进行了系统阐述,分析了机器学习在涡轮机械上述领域应用中的问题及发展趋势。As a widely used mechanical structure in industry,turbomachinery is an indispensable part of the energy field.With the development of turbomachinery,it is dificult to achieve the increasing engineering needs since the improvement of its performance is limited by experiment and numerical simulation techniques.In recent years,artificial intelligence(Al)and data-driven approach have been developed rapidly.Being a main branch of AI,machine learning has attracted much attention and shown great application potential in many fields.The application of machine learning in turbomachinery has the advantages of high efficiency and good generalization ability for the optimization design of turbomachinery components,and is more intelligent and efficient for life prediction,health monitoring and fault diagnosis.It brings new opportunities for the development of turbomachinery.The basic principles of machine learning are outlined and the classical machine learning algorithms are summarized.The application progress of machine learning in turbomachinery components such as turbine blades,turbine disks and sealing components are systematically described.Finally,the problems and development trends of machine learning in turbomachinery are specially analyzed.

关 键 词：机器学习 涡轮机械 叶片 轮盘 密封 

分 类 号：V231.1[航空宇航科学与技术—航空宇航推进理论与工程]