基于人工蜂群算法的大规模武器目标分配研究
Large-scale weapon-target allocation based on an artificial bee colony algorithm
周玉虎 1王桐 1陈立伟 1付李悦 1韦正现2
作者信息
- 1. 哈尔滨工程大学 信息与通信工程学院,黑龙江 哈尔滨 150001
- 2. 中国船舶工业系统工程研究院,北京 100094
- 折叠
摘要
针对大规模武器目标分配问题,本文提出一种改进的多目标武器目标分配模型,该模型将武器平台泛化为武器,并将武器平均飞行时间作为第 2 个优化目标.为有效解决这类问题,本文还提出了改进的自适应离散多目标人工蜂群算法.该算法基于人工蜂群算法和非支配排序策略,引入了自适应算子操作数、重用蜜源探索信息的变异概率策略,并通过蜜源之间、蜜源与外部解集之间的交互以提高算法的收敛性,通过算子的随机选择保持种群多样性.最后通过不同规模武器目标分配的对比实验,证明了所提自适应算子操作数与重用蜜源探索次数的变异概率策略的有效性,并与MOABC、MOPSO、NSGA-II算法在反向世代距离、超体积、时间 3 个方面进行比较,本文算法能够在保证时效性的前提下得到质量更好的Pareto解集.
Abstract
Addressing the problem of large-scale weapon-target allocation,we propose an improved multi-target WTA model that generalizes weapon platforms into weapons and takes the weapon average flight time as the second optimization objective.Herein,to effectively solve such problems,an improved adaptive discrete multi-objective ar-tificial bee colony algorithm is additionally proposed.This algorithm is based on an artificial bee colony algorithm and a non-dominated sorting strategy and introduces adaptive operator operands and a mutation probability strategy that reuses nectar source exploration information.Furthermore,its convergence is improved through the interaction between nectar sources,as well as nectar sources and external solution sets,and the population diversity is main-tained through the random selection of operators.Finally,comparative experiments of weapon-target assignments of different scales proved the effectiveness of the proposed adaptive operator operand and the mutation probability strat-egy of reused nectar source exploration times.The proposed algorithm is compared with MOABC,MOPSO,and NSGA-II algorithms in terms of interuerted generational distance(IGD),hyper volume(HV),and time,and found to obtain Pareto solution sets with better quality on the premise of ensuring timeliness.
关键词
人工蜂群算法/大规模/武器目标分配/多目标优化/自适应/算子操作数/非支配排序Key words
artificial bee colony/large scale/weapon-target assignment/multi-objective optimization/adaptive/operator operands/non-dominant sorting引用本文复制引用
基金项目
国家自然科学基金(61102105)
国防科技重点实验室项目(6142209190107)
先进船舶通信与信息技术工信部重点实验室项目(AMCIT2101-08)
中央高校基本科研业务费专项(3072021CF0813)
出版年
2024
国家科技期刊平台
NSTL
万方数据