Dynamic Tabu Search for Collision Avoidance in Autonomous Maritime Ships

Abstract

This article presents an innovative approach to maritime collision avoidance, featuring a redesigned tabu search algorithm that creates effective solutions across various sea regions by incorporating COLREG, regional rules, ship capability, and good seamanship. The algorithm's dynamic memory adapts solution spaces to changing conditions, ensuring optimal outcomes by modifying tabu, avoiding local minima, and initiating searches in probable solution regions. Notably, the method generates multiple evaluated solutions considering criteria such as risk, implementation cost, legality, and applicability. The proposed comprehensive approach promises to enhance maritime safety and operational efficiency through adaptable collision avoidance strategies.