Correction and prediction of the spacecraft’s orbital motion by artificial neural networks

The problem of transfer between two points of space exploiting the artificial neural networks is considered. It is assumed that the onboard computer of the spacecraft is equipped by the two neural networks: the predictive one and the controlling one. Both the neural networks are trained within the framework of the unperturbed two-body problem model. The purpose of the work is to study the capabilities of the proposed neural networks for the correction of motion in cases of perturbations.