Course Tracking Control for Smart Ships Based on A Deep Deterministic Policy Gradient-based Algorithm

Smart ships have become a research focus due to their wide application prospects. This paper applies the Deep Deterministic Policy Gradient-based (DDPG) algorithm to the course tracking control of smart ships, which improves the global tracking error convergence of the system. In particular, the DDPG-based algorithm uses neural network to approximate the corresponding value function, meanwhile combines deterministic policy gradient. It overcomes the shortcoming of traditional intelligent algorithms, which requires large amounts of training data. Moreover, the online training of the proposed method is capable of addressing the problem of uncertainties in ship motion under specific scenarios. The simulation results show that the DDPG-based algorithm is capable of providing satisfactory results and performs better than the traditional PID-based control algorithm.