Research On Large Envelope Model Identification And Transition State Control Of Turbofan Engine

With the improvement of aircraft and engine performance,the requirements of aeroengine control system design are also increasing,and the control variable is no longer just the fuel flow.Compared with the traditional control model modeling method and variable gain control,the current control system design puts forward higher requirements.Aeroengine state variable modeling is an important part of aeroengine control system design and simulation test,which can reduce test risk and cost.Therefore,based on a certain type of turbofan engine,this article mainly focuses on the full envelope control model modeling and transition state control law of turbofan engine.The main research contents are as followsIn order to solve the problem of turbofan engine state variable modeling,the conventional small disturbance and fitting method are used to obtain the linearized model under different working conditions,thereby obtaining the small deviation linear variable parameter model with high pressure rotor speed.The state variable of the model is only the increment of speed,coordinating with the steady-state base point model to obtain the large deviation state variable model,but this method still has the disadvantage of low precision in a large range.Based on this,combined system identification with least square method,a method of turbofan engine identification modeling is proposed,and the state variable model with the absolute speed as the state variable is obtained.Taking the component level model of a turbofan engine as an example,the mixed signal is proposed as the excitation signal to improve the modeling accuracy and model generalization ability.To study the least square algorithm for turbofan engine system identification further,combining the advantages of steepest descent method and Gauss Newton method,L-M estimation algorithm can be used to obtain the optimal solution of the identification problem.The simulation results show the effectiveness of the modeling method.Under the same flight conditions,considering the problem that the dynamic characteristics change continuously with the speed,the polynomial nonlinear model(PN)is obtained based on the linearized model established by system identification.In view of the large difference of engine performance at different altitudes and Mach numbers in the full flight envelope,based on the polynomial nonlinear model,the timevarying parameters of altitude and Mach number are added,and the nonlinear parameter varying(NPV)is obtained by multiple nonlinear regression method,Finally,a real-time NPV model with time-varying parameters of speed state,altitude and Mach number is obtained,which can better describe the complex dynamic characteristics of turbofan engine with large envelope and large speed range.The simulation results show that the large envelope NPV model meets the requirements of high accuracy.To the design problem of transition state controller aiming at the typical operating point of turbofan engine,based on the polynomial model of turbofan engine,the preset performance theory and the sum of squares(SOS)programming theory are applied to the transition state controller design of polynomial system,and a state dependent design method of preset performance transition state controller is proposed.Based on the component level model of a turbofan engine,the controller is simulated and verified,the results show that the transient state controller can ensure the stability and robustness of the system,the parameters of the dynamic transition process meet the design requirements,and the transition time is short.