Abstract:Automatic collision avoidance decision making for vessels is a critical challenge in the development of autonomous ships and has become a central point of research in the maritime safety domain. Effective and systematic collision avoidance strategies significantly reduce the risk of vessel collisions, ensuring safe navigation. This study develops a multi-vessel automatic collision avoidance decision-making method based on deep reinforcement learning (DRL) and establishes a vessel behavior decision model. When designing the reward function for continuous action spaces, the criteria of the "Convention on the International Regulations for Preventing Collisions at Sea" (COLREGs) were adhered to, taking into account the vessel's collision risk under various encounter situations, real-world navigation practices, and navigational complexities. Furthermore, to enable the algorithm to precisely differentiate between collision avoidance and the navigation resumption phase in varied vessel encounter situations, this paper incorporated "collision avoidance decision making" and "course recovery decision making" as state parameters in the state set design, from which the respective objective functions were defined. To further enhance the algorithm's performance, techniques such as behavior cloning, residual networks, and CPU-GPU dual-core parallel processing modules were integrated. Through simulation experiments in the enhanced Imazu training environment, the practicality of the method, taking into account the effects of wind and ocean currents, was corroborated. The results demonstrate that the proposed algorithm can perform effective collision avoidance decision making in a range of vessel encounter situations, indicating its efficiency and robust generalization capabilities.

COLREG Scenario classification and Compliance Evaluation with temporal and multi-vessel awareness for collision avoidance systems

Grounding hazard considerations in evaluation of COLREGS collision avoidance algorithms

Stochastic COLREGs Evaluation for Safe Navigation under Uncertainty

Collision Avoidance Algorithm Based on COLREGs for Unmanned Surface Vehicle

Ship collision risk analysis: Modeling, visualization and prediction

Multi-Ship Collision Avoidance Decision-Making Based on Collision Risk Index

A Multiple Vehicle Sensing Approach for Collision Avoidance in Progressively Deployed Vehicle Networks.

Codification of Good Seamanship in Complex and Congested Waterways

Towards A COLREGs Compliant Autonomous Surface Vessel in a Constrained Channel

Autonomous Navigation in Confined Waters -- A COLREGs Rule 9 Compliant Framework

Multi-ship collision avoidance decision-making and coordination mechanism in Mixed Navigation Scenarios

A COLREGs-Compliant Deep Reinforcement Learning Approach

COLREGs-Adaptive trajectory planning and decision-making in maritime autonomous surface ships

Exploration of COLREG-relevant parameters from historical AIS-data

A Systematic Approach for Collision Risk Analysis Based on Ais Data

COLREGS-based Path Planning for Ships at Sea Using Velocity Obstacles.

A Two-Stage Dynamic Collision Avoidance Algorithm for Unmanned Surface Vehicles Based on Field Theory and COLREGs

An Improved NSGA-II Algorithm for MASS Autonomous Collision Avoidance under COLREGs

Local Collision Avoidance Algorithm for a Unmanned Surface Vehicle Based on Steering Maneuver Considering COLREGs

COLREG and MASS: Analytical review to identify research trends and gaps in the Development of Autonomous Collision Avoidance

Risk-based implementation of COLREGs for autonomous surface vehicles using deep reinforcement learning