Multi-model Switching Control For Civil Aircraft Recovery From Special Flight

Loss of control in one civil aircraft flight has severely threatened the flight quality and safety.The overdue loss of control situation recovery could cause serious flight accident.The flight fault is one of the major causes in loss of control.Therefore,the prediction of flight fault and recovery research on special state after loss of control fault can improve the reliability of civil aircraft.According to above research background,the work mainly carried out by thesis is as follows.(1)Complex civil aircraft model linearization methodology.Utilize linear parameter varying theory(LPV)to achieve the linearization of complex nonlinear system with close coupling characteristic.Considering the difficulty in complex model transformation to polytypic form,the combination of generation algorithm,which belongs to heuristic search algorithm,and function substitution theory,which relies on grid segmentation LPV theory,are selected to introduce the linear normal flight model,single engine fault model and flight model with wind disturbance.By using time domain response analysis,achievable sets analysis,period model analysis and so on,the thesis proves the LPV model synthesized with generation algorithm is highly efficient;By simulation of aircraft landing procedure when passing through wind field,the thesis proves the LPV model of wind disturbance affection is valid;The simulation of engine fault simulation verifies the fault LPV model is highly approximate to original nonlinear engine fault model.(2)Fault detection and isolation(FDI)system construction on multi model adaptive estimator(MMAE)theory.Design colored noise Kalman filter,residue signal and fault probability calculation module.By increasing passive monitoring mechanism and procedure flag,the simulation proves the FDI system performance can be improved in some degree.(3)Recovery control strategy research and design.Considering passive FDI,active FDI and model form of FDI system,study and simulate two types recovery controller,which are r performance robust control synthesis control and model predictive control(MPC)with control constraint,state constraint and control increment constraint.Comparing to linear quadratic regular(LQR)controller,the controller simulation results illustrate the designed controllers are more effective in reference signal tracking simulation than LQR.(4)Fault tolerant control(FTC)system synthesis and wind disturbance parameter identification.Based on robust synthesize controller belonging to passive control and MPC controller belonging to active control,build FTC system of two common flight fault models,which are single engine fault and elevator fault under wind disturbance.Comparing to FTC system on LQR regular,the simulation indicates the robust synthesis controller is able to restrain wind disturbance;By adopting controlling horizontal stabilizer strategy in elevator fault recovery,the robust synthesis controller FTC system can recover elevator fault and achieve the high precision tracking of reference signal;For recovering single engine fault,the FTC system on MPC,which is to solve control allocation problem towards remaining normal working throttles,shows superior ability to maintain original flight state when the fault occurs.