The nonlinear dynamic model of a supermanenverable aircraft with thrust vector is established, and the process of
modeling the aerodynamic forces, aerodynamic moments and the engine is analyzed. An adaptive inversion control method based on neural network is proposed for supermaneuverable flight. Firstly, the dynamic inversion is applied to design the control laws for the two sub-systems separately. Then the back propagation neural network (BPNN) is used to compensate the parametric uncertainties and the external disturbances. The results of “Cobra”maneuver simulation show that, the designed control law has good performance under the condition of high angle of attack, and ensures the stability of the closed-loop system.