Disturbance observer based reliable H∞ fuzzy attitude tracking control for Mars entry vehicles with actuator failures

This paper introduces a disturbance observer (DO) based reliable H∞ fuzzy attitude tracking control design for Mars entry vehicles with actuator failures. Initially, to reduce the complexity of Takagi-Sugeno (T-S) fuzzy modeling, the two time-scale decomposition technique is used to divide the original nonlinear attitude tracking error model of Mars entry vehicles into a slow subsystem describing the attitude kinematics and a fast subsystem describing the attitude dynamics. The dynamic inversion control (DIC) method is subsequently applied to the slow subsystem to generate the angular velocity command. Then, the T-S fuzzy modeling method is employed to exactly represent the fast subsystem and a disturbance observer (DO) is constructed to estimate the modeled disturbance based on the derived tracking error fuzzy system of angular velocity. By the technique of linear matrix inequalities (LMIs), a DO based reliable H∞ fuzzy controller of attitude tracking is developed to stabilize exponentially the angular velocity tracking error and the modeled-disturbance state estimation error with an H∞ tracking performance both in nominal and actuator failure cases. Furthermore, it is shown that the original nonlinear tracking error system is also exponentially stable and satisfies an H∞ tracking performance both in nominal and actuator failure cases under the proposed fuzzy control law together with the DIC law, provided that the timescale separation between the fast and slow subsystems is valid. Finally, simulation results illustrate the effectiveness of the proposed design method.