粒子群算法在多船避碰决策中的应用

Studies on turning angle to avoid collision between ships with PSO arithmetic

多船避碰是船舶避碰中最复杂的问题,也是船舶自动避碰方法研究中的难点之一。对许多学者的多船避碰研究进行分析,将最近会遇距离DCPA、最近会遇时间TCPA、两船距离、相对距离、本船转向角等作为基本评判参数,利用雷达进行一系列的观测,获得避让要素,建立碰撞危险度的评价模型。应用粒子群算法找出最优的解决方案,得出最优转向避碰幅度解。该方法不仅有助于解决多船会遇情况下的本船最优转向角度值,而且也有助于多船避碰决策系统的智能化设计与开发。

It is the most complicated problem in the collision avoidance of ship that multiple ships encounter with each other at sea, and also is one of the difficult points for ship＇ s automatic collision avoidance. With many scholar＇ s research of avoid multi-ship collision. An evaluation model for collision risk factor is proposed based on the factors such as the distance of closest point of approach, time to the closest point of approach, distance of target ship, relative distance and change point, also using a series of radar observations obtains avoided meeting by the key element. Application PSO arithmetic to find the best option. The simulation result shows that this arithmetic is beneficial to the ship＇ s solution of optimization magnitude and effective. The simulation result shows that this method is beneficial not only to the ship＇ s solution of optimization magnitude of anti-collision turning for the situations of multi-ship encountering with each other but also to the intelligent design and development of ship avoiding-bump decision-making system.