基于“单救援艇-多待救目标”救援场景智能优化算法性能对比研究

Comparative study on the performance of intelligent optimization algorithms in the“single rescue vessel-multiple rescue targets”scenarios

原文传递

导出

摘要目的:探讨海难事故中可被应用于无人救援艇搜救路径生成的3种智能优化算法的性能差异。方法:利用Python 3.9编程工具构建3个分别存在10、30、90名待救人员的救援场景,采用遗传算法、蚁群算法和模拟退火算法3种智能优化方法实现对不同救援场景下救援艇搜救路径的优化,每种算法在每个救援场景下运行10次,输出优化路径距离和优化耗时2个参数,采用SPSS 26软件对其进行统计描述。结果:在10名待救人员场景下,蚁群算法平均优化路径距离最短,为2304.419 m,路径距离方差为0.000 m 2;模拟退火算法平均优化耗时最短,为0.007 s,耗时方差为0.000 s 2。在30名待救人员场景下,蚁群算法平均优化路径距离最短,为4401.945 m,路径距离方差为630.617 m 2;模拟退火算法平均优化耗时最短,为0.464 s,耗时方差为0.001 s 2。在90名待救人员场景下,蚁群算法平均优化路径距离最短,为8102.946 m,路径距离方差为3678.682 m 2;模拟退火算法平均优化耗时最短,为12.779 s,耗时方差为0.434 s 2。结论:综合优化路径距离和优化耗时2个参数分析,本研究背景下蚁群算法优于遗传算法和模拟退火算法。 Objective To explore the performance differences of three intelligent optimization algorithms in generating search and rescue paths for unmanned rescue vessels in maritime disasters.Methods Using Python 3.9,three rescue scenarios with 10,30,and 90 rescue targets were constructed.Genetic algorithm(GA),ant colony optimization(ACO),and simulated annealing(SA)were employed to optimize the rescue paths in each scenario.Each algorithm was run 10 times per scenario,and two parameters—optimized path distance and optimization time—were recorded.SPSS 26 was used for statistical analysis.ResultsIn the 10-target scenario,ACO achieved the shortest average optimized path distance of 2304.419 m with a variance of 0.000 m²,while SA had the shortest average optimization time of 0.007 s with a variance of 0.000 s².In the 30-target scenario,ACO yielded the shortest average optimized path distance of 4401.945 m with a variance of 630.617 m²,and SA recorded the shortest average optimization time of 0.464 s with a variance of 0.001 s².In the 90-target scenario,ACO produced the shortest average optimized path distance of 8102.946 m with a variance of 3678.682 m²,while SA achieved the shortest average optimization time of 12.779 s with a variance of 0.434 s².ConclusionBased on the analysis of optimized path distance and optimization time,ACO outperformed GA and SA in the context of this study.

作者毛君赫李峰李炳锋寿家琛胡家庆 Mao Junhe;Li Feng;Li Bingfeng;Shou Jiachen;Hu Jiaqing(Faculty of Military Health Service,Naval Medical University,Shanghai 200433,China)

机构地区海军军医大学卫生勤务学系

出处《中华航海医学与高气压医学杂志》 2025年第5期503-506,共4页 Chinese Journal of Nautical Medicine and Hyperbaric Medicine

基金军队后勤科研重大项目(AHJ22C003)。

关键词遗传算法蚁群算法模拟退火算法无人救援艇救援路径优化 Genetic algorithm Ant colony optimization Simulated annealing Unmanned rescue vessel Rescue path optimization

分类号 U698.7 [交通运输工程—港口、海岸及近海工程]