Prediction And Research On The Vibration Value Of Low-pressure Rotor Of Civil Aero-engine

The running data of civil aero-engine is an important reference for airlines to make engine maintenance programs,and the vibration fault of engine low-pressure rotor is one of the common faults.Aiming at the difficulty of establishing engine model based on engine principle,the machine learning algorithm and mathematical statistics theory are applied to study engine vibration prediction model based on extreme learning machine improved by particle swarm optimization(PSO-ELM)and engine vibration fault prediction model based on time series analysis method.These studies provide new research ideas for the research of vibration value of civil aero-engine,and provide reference for the prediction of low-pressure rotor vibration fault of airline engine and the formulation of maintenance programs.In machine learning,an extreme learning machine(ELM)model is established for engine low-pressure vibration values and related parameters in this paper.Particle swarm optimization(PSO)algorithm is used to iteratively optimize the model parameters.The average impact value(MIV)algorithm is used to select the input parameters of the model,and the cross validation method is used to select the number of hidden layer neurons.The test data are used to validate and analyze the trained network model.The results show that the PSO-ELM model can achieve ideal accuracy under low hidden layer neurons in aero-engine low-pressure rotor vibration prediction,and the prediction effect is pretty good.In the field of mathematical statistics,the time series analysis method is used to establish autoregressive integral moving average(ARIMA)model for time series composed of regression parameters describing vibration value and N1 speed characteristics of low pressure turbine(LPT)shaft of engine.The ARIMA model is used to predict the trend of regression parameters which characterize the low-pressure vibration-N1 speed characteristics,and according to the overall trend of parameters in expert modeler and ARIMA model,two low-pressure vibration fault symptoms of the engine studied recently are determined.Modeling the normal engine has not found this trend,which verifies the credibility of the model analysis results.