Vessel Trajectory Prediction at Inner Harbor based on Deep Learning using AIS Data

Abstract This study proposes a novel approach for predicting vessel trajectories in the inner harbor of Busan Port using Automatic Identification System (AIS) data and deep learning techniques. Linear interpolation was applied to address unequal time intervals and limited data. Recurrent Neural Network (RNN), Long Short-Term Memory (LSTM), Bidirectional Long Short-Term Memory (Bi-LSTM), Gated Recurrent Unit (GRU), and Bidirectional Gated Recurrent Unit (Bi-GRU) models were trained and evaluated, with LSTM achieving the best performance. The study also identified the critical prediction area for Vessel Traffic Service Operator (VTSO). The proposed method can contribute to enhancing safety and efficiency of vessel traffic management in complex port environments.