Control Synthesis For Integrated Flight/Propulsion System

The key techniques of the integrated flight/propulsion control system (IFPCS) are considered in this dissertation. Hierarchical decentralized control system synthesis is used and studied field contains trajectory optimization, nonlinear dynamic inversion flight control, thrust vectoring control, robust control for turbine engine, high stability control for propulsion system.Artificial neural networks are used to deal with the full nonlinearity which is important while designing the nonlinear dynamic inversion controller to control the flight path angle. A feed-forward neural networks are used to approximate to full nonlinear inverse mapping. According to analyzing the errors of the system, one adaptive neural networks are used to compensate the error of simplified system and the mapping error of feed-forward neural networks. The topology of the feed-forward neural networks and the adaptive laws of the adaptive neural networks are presented also.For fast dynamics with thrust vectoring, special attention is given to the design of a control distribution scheme which switch the thrust vectoring on to improve the maneuverability while the speed is too low. For the trajectory following in low altitude penetration case, proportional guidance plus flight path angle predictive control is presented which can make the aircraft follow the reference trajectory accurately.The principle of direct optimization is introduced. Details of applying this method to flight trajectory generation with constraint of both propulsion system and flight system are presented. Combined with digital terrain map, direct method is applied to the three dimensional feasible trajectory optimization for low altitude penetration, and modified simplex algorithm is used to solve the parameters in optimizing, the trajectory can be optimized real-timely onboard of an aircraft.A high-fidelity mathematic model of nonlinear integrated flight/propulsion system is constructed in this paper. Both the nonlinear model of aircraft and the nonlinear components model of turbofan engine are investigated, and their synthesis method is discussed. Inlet pressure distortion and its influences to characteristics map of turbine engine are considered. The model of converging-diverging nozzle and the flight dynamics caused by thrust vectoring are developed also.Applying LMI based robust control system synthesis to turbine engine is discussed and modeling of small perturbation state variable model (SVM) for design of turbineengines" multivariable control system is studied. The controller for system with parameter uncertainty caused by inlet distortion and thrust vectoring is investigated and validated in the simulation environment.High Stability Engine Control (HISTEC) is introduced also. An approach of prediction of the angle of attack using general prediction (GP) is presented, and this approach can accurately predict the AOA while the aircraft in maneuver. The HISTEC system can tolerate the inlet distortion by reducing pressure ratio of the fan, so the engine can operate under given stability margin adaptively.