基于拟态物理学优化算法的船舶变速避碰决策

Artificial Physics Optimization Algorithm for Collision Avoidance Through Speed Alteration

为能在狭窄水域多船会遇情况下快速给出安全有效的避碰决策,提出一种基于拟态物理学优化算法(Artificial Physics Optimization, APO)的船舶变速避碰决策方法。该方法利用APO算法的快速寻优能力,根据《国际海上避碰规则》对变速行动幅度的要求,限定算法中可行解的空间,考虑变速冲程和冲时对会遇的影响,提高最近会遇距离值的计算精度,建立基于碰撞危险度和变速能量损失的目标函数,迭代进化获得全局范围内最优的速度值,即船舶的决策速度。狭窄水域多船会遇案例的仿真结果表明:决策速度下船舶能够安全有效地避让各目标船舶,该决策方法可行且有效,能够为船舶驾驶人员提供避碰决策支持。

When navigating in narrow waters and encountering multi-ships, speed alteration is the more effective measure to avoid collision than course alteration due to the space constrain. The APO (Artificial Physics Optimization) algorithm is proposed to deal with such situation. The algorithm gives accurate speed alteration suggestions according to the requirements from the International Regulations for Preventing Collisions at Sea (COLREG) and taking into consideration the effect of speed alteration with the transit time and the moving distance during the period counted. The objective function is established based on the risk of collision and the energy loss caused by speed alteration. The algorithm solves the optimal ship speed by iterative evolution. The simulation results show that the decision method is feasible and effective in avoiding target ships, providing decision support in collision avoidance for ship pilots .