Study On Gas Path Fault Diagnosis Technology Of Aeroengine

Aero-engine is the heart of anaircraft.The operation state of aero-enginehas a direct influence on the safety and reliability of the aircraftas well as man's life and property security. So,it is necessary to investigate theaero-engine fault diagnosis system, which is helpful to realize developmentof the maintenance strategy from the traditional periodic maintenance to on-condition maintenance, which is more scientific and economic. Moreover, the study on aero-engine fault diagnosis can improve the security and reliability of the operation. According to the statistics, the gas pathfaultsof the engineaccount for about 90 per cent of all engine failures. From this point of view, this paper mainly concerns the gas path fault diagnosis of the double-spool turbofan civil aviation engine.Based on the small deviation theory, the modeling of gas path fault of aero-engineisintroducedin detail.And the "coefficient matrix" has been deduced, which can easily show the relationship between measurable thermal parameters and performance parameters of aero-engine.The model of gas path fault of aero-enginecan be considered as the theoretical basisfor the condition monitoring and performance prognosis of the aircraft engine. Collecting gas path fault rules of aero-engine, the variations of the measuring parameterscan be obtained by solving the model of the engine when one or more components have faults. The variations of the measuring parameters can bethe judgment principles of the condition monitoring and the fault isolation.The development of artificial intelligence technologyprovides great opportunities to develop fault diagnosis neural network. RBF neural network has beenutilized in areo-engine gas path fault diagnosis. And the simplified RBF neural network is proposed to be used for quantitative analysis system, which has the traits such as simple architecture,fast learning convergence, precise output and the good generalization.According to the steps and processesof fault diagnosis, the principles ofgas path fault diagnosis system of aero-enginewere proposed.The integration of the mathematical modeling and the RBF neural network will make gas path fault diagnosis system morepowerful.