Neural Network and Adaptive Inversion for Re-entry Vehicle Control

This paper firstly analyses the kinetics model of re-entry vehicle, then presents a control method using neural network and adaptive inversion to overcome the flaws existing in other conventional control methods such as proportional-derivative (PD) and recently have extended to a special case of proportional-integral (PI) desired dynamics using implicit model-following. This new controller can adapt to nonlinear and strong couple of the plant, and the controller is not sensitive to the variety of flight condition and other uncertainties. Simulation results for a re-entry flight model are presented to illustrate the good performance of the controller.