Robust Adaptive Control Based On Fuzzy Theory For Aerospace Vehicle

At present, many countries in the world are speeding up their efforts in the research on next generation reusable flight vehicles-aerospace vehicles (ASVs). The United States and European countries have their own aerospace projects and have achieved significant research progress, while, the research on hypersonic vehicles in our country is at in the initial stage. The control system design for the ASVs is a challenge research topic due to their multi-mission profiles, large attitude maneuvers and complicated flight conditions. In this dissertation, four relative problems, i.e. Takagi-Sugeno (T-S) fuzzy modeling and analysis for a conceptual ASV,fuzzy control for uncertain nonlinear system, T-S fuzzy system modeling training and fuzzy adaptive control system design, are studied.First of all, a whole of kinetic equations and motion equations during the reentry phase of an ASV are presented based on the publicated literatures and the contributions of our lab. The coefficients of aerodynamic force and moment are given as functions of angle of attack, Mach number and control surface deflections. The propulsion system is a combination of reaction engine. Rigid-body mass moments of inertia and center of gravity location are time-varying functions of vehicle weight. Open-loop dynamics demonstrate that the proposed model can embody the characteristics of ASV, such as complicated nonlinearity, strong coupling, fast time-varying and uncertainties. The model has a certain representative and can be used to meet the requirements of research and simulation for advanced guidance and control problems.Next, based on the T-S fuzzy approximation theory and the reentry attitude dynamics of ASV, the fuzzy rules, the membership functions and the fuzzy sub-systems are established. A simulation model for the reentry attitude dynamics based on T-S fuzzy system is presented, and it lays a foundation for the following controller design and analysis.Then, based on the presented T-S fuzzy model for ASV reentry attitude dynamics, a new design method of T-S fuzzy guaranteed cost control law with pole constraints is proposed under the guidance of regional pole placement theory. The external disturbance is restricted by using H∞control technique, and a new fuzzy robust control design method is presented. The Lyapunov stability theory is used to prove asymptotic stability of all the states of the closed-loop system. With the different requirement of performance index, state feedback and output feedback controller for ASV attitude dynamic system are designed respectively. The next, a new design technique of fuzzy robust tracking control based on fuzzy feedforward is proposed. By taking the advantages of T-S fuzzy systems, feedforward control scheme for linear system is extended to nonlinear system. Without the need of the augmented system of linear matrix inequality (LMI) based T-S fuzzy tracking control method, the dimension of system design is reduced, and less conservative is achieved. The asymptotic stability of system tracking error is guaranteed by using Lyapunov stability theory. Considering the system uncertainties and external disturbance, a fuzzy robust tracking controller of the controlled system is designed. The simulation results for the output tracking of complex nonlinear system and attitude angles tracking of ASV demonstrate the effectiveness of the proposed method.Then, to enhance the adaptation of T-S fuzzy model as system parameters changes, a new Levenberg-Marquardt (L-M) algorithm is proposed, which can be used for Takagi-Sugeno fuzzy modeling. Based on standard L-M algorithm and under the condition of local error bound, the iteration step is defined as a function of approximation error according the different values of iteration parameters. A parameter iteration formula of the modified L-M algorithm is given, and its quadratic convergence is analysed theoretically. Then it is applied to training the ASV attitude dynamic systems based on T-S fuzzy model. This method can adjust the parameters of each linear polynomials and fuzzy membership functions on line. In addition, it does not rely on expert experience excessively when using it for systems modeling. Compared the proposed method with the standard L-M method, the convergence speed is accelerated and high approximation precision is achieved.Finally, a new adaptive regulator is studied. It can adjust the parameters of fuzzy adaptive controller. Compared with gradient adaptive law, the parameter convergence rate is accelerated obviously. Aimed at the problem of output tracking control for complex nonlinear system, a stable indirect adaptive controller is presented and the system oscillations due to the low convergence speed of parameters are effectively improved. Then the control scheme is extended to multi-input multi-output (MIMO) square nonlinear plants with non-modeled dynamics and an attitude indirect fuzzy adaptive controller for ASV is designed. In addition, considering the effect of external disturbance, a robust control term is introduced and the control effect is further improved.