Linear-parameter-varying Control Design For Aero-Engine

The aeroengine control system generally uses the gain-scheduled control method. However, with the increasing requirements of aeroengine performance, the traditional gain-scheduled control method cannot ensure the global stability of the system when the parameter is changing in a wide range. Meanwhile, the design cycle of the traditional method is too long. The robust gain-scheduled control based on the linear-parameter-varying(LPV) system for turbofan engine is researched in this thesis.Two methods are applied to deal with the modeling problem in the robust gain-scheduled control, i.e. frequency response method and small disturbance method combined with least square method. The transfer function model and state variable model are built by using these two methods. Based on the linearization model, LPV model is built by polynomial fitting. As for the LPV model of transfer function, a kind of PI control method for high pressure rotor speed is presented, and its effectiveness is validated by simulation analysis.In view of the rotor speed changing in a wide range and the realization problem of the controller for aeroengine, through the study in the robust stability of transfer function model and combining the LPV control theory with PI control structure, a robust LPV/PI control method based on the polynomial sum of squares programming(SOSP) method is presented. In order to test the control effect, the controller is applied on the component-level model of turbofan engine to be simulated and researched.Based on the research in the robust stability of state variable model, a method of gain-scheduled control is put forward to the high pressure rotor speed controller design. Combining with a gain-scheduled control structure, which can ensure the system with no static error, the robust stability constraint conditions are obtained on the basis of the bounded real lemma and sum of squares theorem. The problem of the controller design can be translated into SOSP problem. By applying the controller on the component-level model of turbofan engine, the effectiveness of the controller is evaluated.