Research On Robust Gain Schedling Control Of Aircraft Engine

With the improvement of advanced fighter and the development of aero-engine design and manufacturing technology,aero-engine control system has gradually developed into a complex multivariable control system.In order to make the engine perform as best as possible under the condition of safety and stability,this dissertation focuses on the multivariable robust variable gain scheduling control method,and studies the linear parameter varying（LPV）modeling,multivariable output tracking control,large-scale robust control and transition control.Aiming at dealing with the discontinuity of linear model coefficients and eigenvalues caused by interpolation of characteristic map,an improved calculation method of component characteristic map based on 3-order spline interpolation is proposed,which improves the continuity of engine linear model coefficients at adjacent working points and thereby improves the accuracy of the model.In addition,a fast model based on LPV model and component flow path calculation is designed to improve the real-time performance.In this model,LPV model is used to replace the iteration of the equilibrium equation,which avoids iterative calculation and improves the real-time performance while ensuring the accuracy of the model.The augmented linear quadratic regulator（ALQR）can’t achieve the optimal control performance under time-varying command.To solve this problem,an augmented linear quadratic tracker（ALQT）based on real-time compensator is proposed accrding to the idea of linear quadratic tracker（LQT）,which improves the output tracking performance.Since it is difficult for ALQR/ALQT to realize overshoot suppression under step reference,an extended augmented quadratic tracker（EX-ALQT）is proposed,and two reference governors,namely reference governor and S-type nonlinear reference governor,are designed based on ALQT to restrain overshoot.To deal with the tracking control with uncertainties,an H₂/H_∞controller is proposed based on game algebra Raccati equation（GARE）.Meanwhile,a real-time performance compensator is designed to improve the tracking performance of the system.In addition,the idea of the compensator in ALQT is applied to GARE,and a real-time compensator is proposed to improve the output tracking performance of the system.Aiming at dealing with the problem of poor control quality of single controller when operating in large-scale points,a large-scale robust LPV control method of H₂/H_∞based on similarity theory is proposed,which realizes the full envelope control of engine.To deal with the problem that the control performance of the controller is degraded due to the large variation of engine state caused by afterburner ignition and flameout,a robust smooth switching method is proposed to ensure the switching stability.The large-scale controllers can be smoothly switched to the controller with high robustness before afterburner ignition and flameout,and thereby the influence on the output is reduced.To deal with the control instability caused by the switching of the working mode of the nozzle during the acceleration and deceleration,a tracking trajectory design method based on the reference governor is proposed,which drives the output trajectory of the engine in a safe area and ensuring the safety and stability during acceleration and deceleration.On this basis,a two-section tracking method is proposed.In the low-speed section,when the nozzle is near the critical state,the tracking trajectory is adjusted to make the engine state and controller parameters change slowly;In the high-speed section,the engine state changes rapidly by adjusting the tracking trajectory,thus ensuring the whole acceleration and deceleration process to be fast nd stable.Finally,according to the existing software and hardware platforms,a real-time simulation and verification platform for aeroengine wide-range controllers is established by using the rapid prototyping method,and the overall structure of the platform and the interaction relationship of each working module are designed.The effectiveness of the control method is verified by real-time simulation experiments.