Research on Ship Trajectory Prediction Using LSTM and BP Based on AIS Data

The rapid development of water transportation has brought certain challenges to the management of ships. In order to assist the management of the Maritime Safety Administration and ensure the safety of ships, this paper compares the role of two prediction algorithms in trajectory prediction based on Automatic Identification Systems (AIS) data: BP algorithm and LSTM algorithm. First, this paper preprocesses the original AIS data and then uses orthogonal optimization to select parameters and conducts the experiments to compare the two algorithms mentioned above. Finally, the distribution map of absolute error, mean absolute error (MAE) and max absolute error table are shown. The results show that the LSTM algorithm is better than the BP algorithm in both overall and detail performance.