发动机及控制系统状态监测方法研究  

Research on Monitoring Methods for Engine and Control System Status

作　　者：陈前景 金鹏[2] 王亚伟 彭瑞轩 鲁峰[2] CHEN Qianjing;JIN Peng;WANG Yawei;PENG Ruixuan;LU Feng(China Aeroengine Research Institute,Beijing 101399;State Key Laboratory of Mechanics and Control for Aerospace Structural,Nanjing University of Aeronautics and Astronautics,Nanjing Jiangsu 210016)

机构地区：[1]中国航空发动机研究院,北京101399 [2]南京航空航天大学航空航天结构力学及控制全国重点实验室,江苏南京210016

出　　处：《软件》2023年第7期13-17,共5页Software

基　　金：航空发动机及燃气轮机基础科学中心项目(P2022-B-V-002-001)。

摘　　要：本文针对发动机及控制系统并发故障的情况,设计一种基于决策级融合的单传感器自调整预测模型,提出一种基于数据驱动算法Meta-OSELM,用于构造发动机传感器的解析余度。并考虑多传感器发生故障时,基于信号融合传感器解析余度算法失效问题,设计一种基于决策级融合的自调整预测模型,实现涡扇发动机的传感器自调整预测,并进行发动机及控制系统并发故障下的故障监测与诊断验证。This article designs a single sensor self tuning prediction model based on decision level fusion for concurrent faults in engines and control systems,and proposes a data-driven algorithm Meta-OSELM for constructing analytical redundancy of engine sensors.And considering the problem of signal fusion sensor analysis redundancy algorithm failure when multiple sensors fail,a self adjusting prediction model based on decision level fusion is designed to achieve sensor self adjusting prediction for turbofan engines,and fault monitoring and diagnosis verification under concurrent faults in the engine and control system are carried out.

关 键 词：发动机控制系统 故障监测 解析余度 Meta-OSELM 

分 类 号：TK418[动力工程及工程热物理—动力机械及工程]