Discrete control of longitudinal dynamics for hypersonic flight vehicle using neural networks

The discrete controller is developed for the longitudinal dynamics of the hypersonic flight vehicle with neural networks （NNs）. The focus is on the control of the Euler equations of attitude dynamics. With back-stepping design, for each step the virtual control is constructed with NNs approximating the unknown dynamics. The design uses less online adaption parameter learning scheme to achieve the control goal. It is guaranteed that the errors of all the signals in the system are uniformly ultimately bounded. The proposed method is verified by simulation of winged-cone model.