基于AIS数据与机器学习的海上补给行为检测

he detection of replenishment behaviors at sea is of great significance for maintaining maritime security and combating illegal activities. The main objective of this paper is to detect replenishment behaviors at sea using Automatic Identification System (AIS) data and machine learning algorithms. A dataset containing 450 ship trajectories of replenishment behaviors was constructed. Methods for extracting various features from trajectories, including time intervals, distances, angles, etc., were proposed. Five classification models were compared in experiments, and random forest model achieved the best classification performance with a recall of 0.960. To verify the effectiveness of the method, a real-time monitoring system for replenishment behaviors was built based on the trained random forest model. From January to November 2023, 781 replenishment eDetection of Replenishment at Sea Based on AIS Data and Machine Learningvents of 59 replenishment ships from 15 countries worldwide were detected, validating the effectiveness of the method. This paper provides a reference for effectively monitoring replenishment behaviors at sea using AIS trajectory data.