| Popis: |
In this paper, we investigate a roll-pitch-yaw integrated autopilot for a short-range air-to-air missile (SRAAM) in the presence of uncertainties in total aerodynamic forces and moments. To design an adaptive controller for command following of angle of attack, sideslip angle, and roll angle, adaptive dynamic surface control (DSC) is developed using neural network. The explosion of complexity in conventional backstepping design is avoided by using DSC. Uncertain nonlinearity is represented by neural network under universal approximation. To achieve adaptive performance, the neural network is learned by updating the weight matrices and gains on-line through adaptation rules that are derived from Lypunov stability theory. It is shown that the proposed control design can guarantee the uniformly ultimate boundedness of all the signals in the closed-loop system, and make the tracking error arbitrarily small. The six-degree of freedom nonlinear missile simulation results validate the feasibility of the proposed control law. |
