Research On Independent Intelligent Fault Tolerant Control For Fighter

This thesis takes a certain model of Fighters as the plant. The purpose is to study the intelligent fault tolerant flight control technology in terms of engineering.Firstly, a gain scheduled tracking control method based on polytopic LPV model and RBF NN is proposed. Approximately the polytopic LPV model is adopted for modeling of the known nonlinear system. Then sufficient conditions of the existence of state feedback controller which makes the LPV system satisfy the requirement of pole location and H∞performance are obtained. According to LMI technology and quadratic separator, the sufficient conditions are transformed to a family of LMIs. Full adaptive RBF neural network control is used to improve the scheme of the LPV tracking control, in order to press close to engineering practice. The effect of the error caused by polytopic LPV modeling is overcome by the neural network.Secondly, a gain scheduled reliable tracking control method based on polytopic LPV model and RBF NN is proposed. Possible damnifications of actuator are considered, and a continuous fault model is used to denote it. A closed loop polytopic LPV system containing failures is established. Reliable state feedback control law which makes the system, in despite of normal or not, satisfy pole placement and H∞requirement, is designed. RBF NN is also used to overcome the modeling error online, in order to achieve satisfactory control purposes.Finally, an intelligent gain scheduled fault tolerant tracking control method based on HLPV model and NN is proposed.Vertex-fault modeling method is used to express faults. Normal and fault system are unified to a HLPV system. Using HLPV theory, sufficient conditions of the existence of state feedback controller which makes the system achieve the requirement of H∞performance and pole location, can be gained. Solving the LMIs, then controller parameters of veteces are available, and scheduled by a BP NN according to the situation of the system. So the system can achieve anticipant control purposes in despite of normal or not.Results of the flight simulation test have demonstrated that the methods in this thesis are all effective.