融合先验信息的民航发动机自适应气路故障诊断方法  

Adaptive gas path fault diagnosis method of civil aviation engine fusing prior information

作　　者：郭庆[1] 刘晓阳 樊俊峰 付宇[1] 左洪福[3] GUO Qing;LIU Xiaoyang;FAN Junfeng;FU Yu;ZUO Hongfu(College of Aeronautical Engineering,Civil Aviation University of China,Tianjin 300300,China;Power Plant Department,Southern Airlines Engineering Technology Branch,Guangzhou 510890,China;College of Civil Aviation,Nanjing University of Aeronautics and Astronautics,Nanjing 210016,China)

机构地区：[1]中国民航大学航空工程学院,天津300300 [2]中国南方航空股份有限公司工程技术分公司动力装置部,广州510890 [3]南京航空航天大学民航学院,南京210016

出　　处：《航空学报》2025年第4期150-161,共12页Acta Aeronautica et Astronautica Sinica

基　　金：国家自然科学基金(U2133202);中国民航大学中央高校基本科研项目(3122022046)。

摘　　要：针对民用航空发动机传感器数量不足、难以实现有效故障诊断的问题,提出了融合先验信息的自适应气路分析模型(AGPAM-PI),提升了发动机故障诊断的准确性与效率。AGPAM-PI结合发动机指印图的先验信息和非线性气路故障诊断方法,首先通过指印图对故障信息进行有效补充,对故障因子求解范围约束限定,然后再利用非线性气路分析模型进行故障诊断。以一台在实际运营过程中发生故障的CFM56-7B发动机数据进行验证,结果表明模型对该发动机的故障进行了准确的定位并通过发动机故障诊断规则分析了发动机故障模式,证明了模型的有效性。相比传统的气路故障诊断方法,先验故障信息的引入增强了单元体故障的识别能力,提高了诊断精度。To address the problem of difficulty in effective fault diagnosis for civil aviation engines due to insufficient numbers of sensors,an Adaptive Gas Path Analysis Model incorporated with Prior Information(AGPAM-PI)is pro⁃posed to enhance the accuracy and efficiency of engine fault diagnosis.The AGPAM-PI combines the prior information of the engine fingerprint and the nonlinear gas path fault diagnosis method.The fault information is firstly supple⁃mented through the fingerprint diagram,the solution range of fault factors is constrained,and then the nonlinear gas path analysis model is used for fault diagnosis.The model is validated using the data from a CFM56-7B engine that experienced a fault during actual operation.The results show that the model accurately locates the fault of the engine and analyzes the engine fault mode through engine fault diagnosis rules,which proves the validity of the model.Com⁃pared with traditional gas path fault diagnosis methods,the introduction of prior fault information improves the ability to distinguish unit faults and enhances diagnostic accuracy.

关 键 词：民航发动机 故障诊断 部件级建模 信息融合 非线性气路分析 

分 类 号：V263.6[航空宇航科学与技术—航空宇航制造工程]