Feasibility of Deep Reinforcement Learning for the real-time attitude control of a satellite system | Publicación