Nearspace vehicle control based on RBFNN

With the assumption that NSV suffers the violent changes of aerodynamic paameters and the outside disturbance in hypersonic condition, we present the integrator backstepping approach based on fully tuned radial basis function neural network (FTRBFNN). The strict proof of the approach´s stability is provided simutanously. The performance analysis for the approach demonstrate that the FTRBFNN has better ability of restaining disturbance than RBFNN, and the integrator term in backstepping approach eliminate the static traking error efficiently.