Research On Robust Control Method For Uncertain Nonlinear Integrated Fire/Flight/Propulsion System

The new generation fighter and its integrated fire/flight/propulsion control system are rapidly developed by the western countries and Russia for scrambling the mastery of the sky. To accomplish the complicated missions safely and effectively, it must have an advanced integrated fire/flight/propulsion system for the new generation fighter. Synchronously, the integrated fire/flight/propulsion system is a complicated, uncertain and multi-variable nonlinear large-scale system. So its control problem is an active subject in the modern control area.In this dissertation, the control methods of MIMO uncertain nonlinear system are designed and used to solve the key control problems involving the integrated iire/flight/propulsion system. On the basis of designing robust flight control system of the new generation fighter, this dissertation is devoted to the design of the integrated fire/flight control system and the integrated flight/propulsion control system and the integrated fire/flight/propulsion control system.First of all, it is devoted to discuss the open-loop characteristics and the nonlinear bifurcation analysis and the maneuver ability for the super-maneuverable flight nonlinear model of the new generation fighter. Then a flight control system is designed in analytical dynamical inversion method based on the time-scale separation principle. The idea of on-line compensating with neural network is introduced to eliminate the inversion error which resulted from the inaccurate fighter's model and the faults of the fighter. The dynamic performance is improved by the weight value's adaptive adjust of the neural network. In the process of adaptive control, to compensate the bad effects of actuator unmodeled dynamic and to improve the system's tracking performance, a pseudo-control hedging based control structure is given, at the same time the convergence of all the signals in the closed system is guaranteed. Then a discussion is devoted to the design of a robust flight control system which is used to overcome the disadvantage of the neural network based on disturbance observer.Secondly, a discussion is devoted to a robust adaptive H control method based on wavelet neural network for a class of MIMO nonlinear uncertain systems and applied this method to design the robust adaptive flight control system. This designed method is valid to the flight control system with unknown model. But the designed robust control method needs a lot of trained data of neural network, so the operation speed is limited. The uncertainty of the system can be regard as the disturbance and it can be approximated by wavelet neural network disturbance observer. Then a robust adaptive H control method and a model predictive control method based on disturbance observer are proposed for a class of MIMO nonlinear uncertain systems, at the same time these methods are used to design the robust adaptive flight control system. The designed robust flight control systemcan not only guarantee stability robust of flight, but also guarantee the flight control system have some robust performance indexes. At the same time, the performance of the flight control system is easily analyzed.Thirdly, a discussion is devoted to study the form principle and structure design for the integrated fire/flight control system of the new generation fighter. The genetic algorithm and variable word collection fuzzy control method are introduced to design the intelligent fire/flight coupler and to improve the performance of integrated fire/flight control system. At the same time, H∞ guidance law of the air-air missile is designed with L2 gain theory and the H2/ Hx mixed control theory to minish the sensitivity to maneuverable goal. The designed robust guidance laws are used to the simulation of air-air missile's attack process. The designed robust guidance laws are more valid than the proportional navigation guidance law.Fourthly, a discussion is devoted to study the integrated flight/propulsion control system. Dynamic identification method based on fuzzy-ne...