.lqg / Ltr Method In The Turbofan Engine Control Applications

Multivariable robust control for a turbofan aero-engine using the LQG/LTR(Linear Quadratic Gaussian with Loop Transfer Recovery) design methodology is considered. For such a complicated plant model, classic feedback control theory can't guarantee it's optimal performance. In this paper, The LQG/LTR methodology provides an integrated frequency-domain and state-space approach for design of BIBO (Binary inputs Binary outputs) control system for the augmented plant model of this kind of engine.In the start stage, working points linear state space models for this turbofan aero-engine have been derived using the Stone method. And the importance of proper scaling the plant to achieve the desired Target- Feedback-Loop(TFL) is emphasized. Next the LQG/LTR procedure is used to design a series of linear control systems at some operating condition of the engine. LQG/LTR at the plant output proceeds in two steps: 1) Design a Kalman filter with desirable performance specifications; 2) Design a sequence of LQ-regulators to make the singular values curve of the open-loop transfer function GK approximate the curve of TFL.The simulation results have illustrated that the command following, noise-suppressing and disturbance-rejection specifications as well as stability-robustness specifications of the control system can meet the requirements. And in predefined operating envelop, a neutral network was used to approximate the control parameters of the LQG/LTR controllers. Furthermore, elementary study was applied to this turbofan engine transient state control, especially acceleration control process.So we can draw such a conclusion: the aero-engine control system is robust so as to meet the control specifications well.