Convention on the International Regulations for Preventing Collisions at Sea,1972(COLREGs) stipulates the vessels’ obligations when they meet, but in the actual navigational environment, the captains don't necessarily believe they have the responsibility to give way according to subjective judgement. Therefore, this study proposes a model for the determination of the give-way vessels, combining vessel characteristics, using asymmetric Gaussian to construct the potential field of stand-on vessels under the perspective of give-way vessels, constructing the cost function based on field theory, selecting the vessel with the lowest cost as the give-way vessel, and carrying out the example verification, to objectively coordinate the avoidance action between the vessels.
With the promotion of The Belt and Road Initiative, waterway transportation has become an important link in international trade. Vessels gradually to high-speed, unmanned development, and the increasing scale of vessels complicates the navigation environment, maritime collision accidents often occur. In addition, according to the survey, more than 80% of vessel collision accidents are caused by human errors, mainly manifested in the captain's misjudgment of whether there is a responsibility to give way (Zhenxing Qiao et al,2021). Therefore, it is crucial to reduce the casualties caused by misjudgment and realize autonomous collision avoidance.
For this reason, many scholars have carried out research on autonomous collision avoidance of vessels, which mainly includes three categories: first, collision avoidance methods based on analytical models, such as Velocity Obstacle (VO) and Artificial Potential Field (APF); second, intelligent algorithms, including Simulated Annealing (SA), Genetic Algorithm (GA); and third, machine learning methods, such as K-Means, Deep Reinforcement Learning (DRL).
In addition, Convention on the International Regulations for Preventing Collisions at Sea, 1972 (COLREGs) stipulate that a vessel on another vessel's starboard in the crossing situation is a give-way vessel, and should give way to another vessel. At present, most of the studies are also under the COLREGs to achieve automatic collision avoidance of give-way vessels, the main idea is to establish a penalty mechanism for violating the COLREGs, integrate the APF (Lingyu Li et al,2021), DRL (Do-Hyun Chun et al, 2021), and Dynamic Window Algorithm (DWA) (Xiaojie Sun, 2023) with the COLRGEs or set up prerequisites to comply with the COLRGEs, as (Ahmed et al,2021) to comply with the COLRGEs take into account the encounter situation and relative angle when calculating the collision hazard, (Haiqing Shen,2018) first converted the COLRGEs into navigable regions and then used the A* algorithm to achieve autonomous navigation of unmanned vessels, (Tam et al, 2013) first determined the priority based on the COLRGEs and vessel maneuverability, normalized the avoidance angle to 30° and determined the linear extension, (Sang-Min Lee et al, 2004) transformed the COLRGEs into mathematical formulas and used fuzzy logic to formulate a collision avoidance path that complied with the COLRGEs.
