Trajectory prediction is an important support for analysing the vessel motion behaviour, judging the vessel traffic risk and collision avoidance route planning of intelligent ships. To improve the accuracy of trajectory prediction in complex situations, a Generative Adversarial Network with Attention Module and Interaction Module (GAN-AI) is proposed to predict the trajectories of multiple vessels. Firstly, GAN-AI can infer all vessels’ future trajectories simultaneously when in the same local area. Secondly, GAN-AI is based on adversarial architecture and trained by competition for better convergence. Thirdly, an interactive module is designed to extract the group motion features of the multiple vessels, to achieve better performance at the ship encounter situations. GAN-AI has been tested on the historical trajectory data of Zhoushan port in China; the experimental results show that the GAN-AI model improves the prediction accuracy by 20%, 24% and 72% compared with sequence to sequence (seq2seq), plain GAN, and the Kalman model. It is of great significance to improve the safety management level of the vessel traffic service system and judge the degree of ship traffic risk.

In this paper, we propose a novel dynamic fuzzy ship domain that considers factors associated with both one's own ship and other ships. This is in contrast to existing ship domain models that operate from the perspective of one's own ship, considering only the factors relevant to that ship. First, the domain was determined by considering the distance of the ships around one's own ship in different directions, which sufficiently accounts for factors associated with one's own ship and with the other ships. At the same time, the factors were chosen based on an analysis of their importance. Second, the domain was dynamic and modelled by establishing the relationship between the domain size and the chosen factors in different directions from one's own ship, obtained by using neural networking and wavelet decomposition. Third, the domain was developed using fuzzy sets related to different safety levels and this related the model to the practical applications of estimating spatial collision risk. The model was calibrated using Automatic Identification System (AIS) data of vessel movements in the Bohai Sea and the northern Yellow Sea. The reasonableness and the superiority of establishing a ship domain model considering the factors affecting both one's own ship and other ships were analysed, and the results show that the new model can determine the spatial collision risk of the navigational situation and is thus suitable for use as part of a ship collision avoidance system.