Dynamic Data-Driven Ship Motion Simulation Toward Visual-Aided Navigation on Water

Visual aids for navigation systems are a key function of intelligent onboard navigation units for ships. These systems visualize the planning decisions of an intelligent navigation unit for computer-aided visual navigation to reduce the cognitive workload of the driving control personnel of an intelligent ship. In this study, we propose a dynamic data driven method to simulate the motion of ships on the water. We constructed a validation dataset for ship motion simulation toward visual-aided navigation considering the combined effects of wind, waves and currents. According to the validation results on this dataset, the proposed simulation method exhibited good accuracy and real-time performance. Under the given control and environmental input series, the mean errors of simulated heading and simulated trajectory were 5.7°/h and 282.6 m/h, respectively; the simulation time was 475 ms for each hour of physical duration. Overall, the proposed simulation method satisfies the need for visual navigation aids and could improve the situational awareness of the driving control personnel of the intelligent ship.