Path Following Optimization for an Underactuated USV Using Smoothly-Convergent Deep Reinforcement Learning | Publicación