【摘要】 针对开阔水域多船会遇场景下的船舶避碰决策问题，提出一种改进后的多目标麻雀搜索算法(Improved Multi-objective Sparrow Search Algorithm, IMOSSA).为增强种群多样性，引入模拟退火算法中的退火策略更新麻雀种群位置；基于精英选择策略，对所有非支配解的拥挤度距离进行降序排列，剔除最小的解并重复直到保留前沿解个数，以提高算法的全局收敛性能，更快去除劣解.仿真结果表明，改进后的IMOSSA在ZDT1上的反世代距离和空间评价指标较MOSSA分别下降了94.5%和89.7%,在多船会遇局面中能够得到符合条件的最优避碰转向角，有助于提高船舶智能避碰决策水平.更多还原

【Abstract】 Aiming at the decision problem of ship collision avoidance in the scenario of multi-ship encounter in open water, an improved multi-objective sparrow search algorithm is proposed. Introduction of an annealing strategy in the simulated annealing algorithm to update the sparrow population location to enhance population diversity. Based on the elite selection strategy, the crowd distance of all non-dominant solutions is sorted in descending order, the smallest solution is eliminated and repeated until the number of leading solutions is retained, so as to improve the global convergence performance of the algorithm and remove the inferior solution faster. The simulation results show that compared with MOSSA,the IGD and SP indexes of IMOSSA on ZDT1 decreased by 94.5% and 89.7%,and can obtain the optimal collision avoidance steering angle that meets the conditions in the multi-ship encounter situation, which is conducive to improve the intelligent collision avoidance decision-making level of ships.更多还原