Disturbance-Observer-Based Robust Flight Control for Hypersonic Vehicles Using Neural Networks

In this paper, robust flight control schemes are proposed for the hypersonic vehicle with uncertainties and external disturbances using disturbance observer and neural networks. The nonlinear mathematic model of hypersonic vehicles is firstly transformed into a multivariable linear system with parametric uncertainty, function uncertainty and external disturbance. Considering the system disturbance generated by an external system, the disturbance observer is given to estimate the external disturbance. On the other hand, neural networks are introduced to approximate system function uncertainties. Combing disturbance observer with neural network, the robust flight control scheme is proposed for hypersonic vehicles. To explicitly tackle the control gain constraint problem, the constrained robust flight control is accordingly developed by combing disturbance observer with neural network. In both proposed robust flight control techniques for the hypersonic vehicle, the gain matrix of disturbance observer and the control gain matrix are solved via Linear Matrix Inequality (LMI). Simulation results are presented to illustrate the effectiveness of the proposed robust flight control schemes for the hypersonic vehicle.