A hypersonic intelligent dynamic inversion control method is proposed to solve the problem that flight-thrust integration of hypersonic vehicle is strongly coupled, the flight environment is strongly uncertain, which leads to the decrease of control quality and the difficulty of parameter design. This method is based on the nonlinear dynamic inversion control (NDI) method. Considering the difficulty of designing control parameters, BP neural network is introduced to optimize and correct the inner and outer loop control gains of NDI online. Furthermore, an error observer is designed to observe and compensate external disturbance to eliminate the influence of flight environment uncertainty on control quality. The simulation results show that the hypersonic intelligent dynamic inversion control method (BP-NDI) can achieve fast and accurate tracking of the pitch angle guidance command, and has good robustness. Compared with the traditional NDI control method, the response speed is faster, the adjustment time is reduced by more than 20%, and the control quality of the hypersonic vehicle is significantly improved.