Decentralized cooperative collision avoidance strategies for conventional and intelligent ships in mixed scenarios
Abstract
With the advancement of the shipping industry and artificial intelligence technologies, the application of intelligent ships has become increasingly widespread. However, the coexistence of intelligent ships and conventional ships brings new challenges to collision-avoidance decision-making. This study proposes a decentralized cooperative collision-avoidance method that integrates the International Regulations for Preventing Collisions at Sea (COLREGs), fuzzy logic inference, and risk-cost assessment to address collision-avoidance problems in mixed navigation environments. The method employs fuzzy collision avoidance angle algorithm to characterize the decision-making behavior of conventional ships, while optimizing the avoidance strategies of intelligent ships through a comprehensive cost function, thereby enabling cooperative avoidance under a distributed framework. Simulation results of typical encounter scenarios demonstrate that the proposed method can accurately identify collision risks and generate appropriate avoidance maneuvers, effectively improving safety and decision-making efficiency in mixed navigation environments. This research provides a feasible pathway for cooperative collision avoidance of intelligent ships and offers valuable insights for future maritime traffic safety management.