基于粒子群-遗传优化算法的船舶避碰决策

为能在开阔水域中提升船舶驾驶员在多船会遇场景下的避碰决策能力,按照国际海上避碰规则(Convention on the International Regulations for Prerenting Collisions at Sea,COLREGs)的要求,综合考虑船舶航行的安全性与经济性,提出一种基于粒子群-遗传(Partide Swam Optimization-Genetic Algorithm,PSO-GA)的混合优化避碰决策算法。基于最近会遇距离(Distance of Close Point of Approaching,dCPA)和最近会遇时间(Time to Close Point of Approaching,tCPA)确定船舶碰撞危险度(Collision Risk Index,ICR)的计算方法,基于转向幅度与航行时间建立避碰决策目标函数。基于PSO-GA算法具有提高收敛精度和加速全局寻优的特点,当ICR≥0.5时,启动PSO-GA算法,获得让路船舶在全局范围内的最佳转向幅度和在新航向上的航行时间。仿真结果表明:与单独使用PSO或GA算法相比,PSO-GA算法能够以较少的迭代次数找到安全经济避碰航线。提出的避碰决策算法能够为船舶驾驶人员在避碰决策中提供参考,有助于提升船舶航行的安全性和降低船舶碰撞事故发生的风险。

Collision Avoidance Decision-Making Based on Particle Swarm Optimization and Genetic Algorithm

The PSO-GA(Partide Swan Optimization-Gentic Algorithm)hybrid optimization algorithm for ship intelligent collision avoidance decision-making in open waters gives proposals that are considered the best for safety and economy under the rules of the COLREGs(Convention on the International Regulations for Preventing Collisions at Sea).The collision risk under evaluation is defined by ICR(Collision Risk Index)based on the dCPA(Distance to Closest Point of Approach)and tCPA(Time to Closest Point of Approach).The objective function is constructed to represent the steering angle for course changing and the sailing time on the new course.When ICR≥0.5,the PSO-GA algorithm is activated to search for the optimal steering amplitude and the corresponding sailing time from the solution space.The PSO-GA algorithm features good convergence speed,high precision and stability.Simulation results on multi-ship scenarios indicate that the PSO-GA can find the solution for the give-way ships with less iterations than sole PSO or GA algorithm,and the suggested maneuvering plans guarantee the safe encountering of ships in an economical way.