Adaptive control of attitude and momentum for space station based on RBF neural networks

Mathematical model of the space station is often assumed to be known exactly when the attitude/momentum controller is designed. However, exact mathematical model is not available due to the constructions or routine operations of the space station. Therefore, a radial basis function (RBF) neural network is adopted to approximate the nonlinear station dynamics and a novel adaptive controller is proposed for the attitude and momentum of the space station. Since the RBF networks can approach any nonlinear continuous functions with arbitrary degree of accuracy, this controller can attenuate the model uncertainties effectively. And also, this controller can establish a proper tradeoff between station pointing and momentum management of the control moment gyroscopes, while satisfying the specific mission requirements. Simulation results of a certain space station indicate that the controller presented above is feasible.