基于进化多目标优化的作战目标分群算法  

Operational Target Grouping Algorithm Based on Evolutionary Multi-objective Optimization

作　　者：赵文栋 张明智[2] ZHAO Wen-dong;ZHANG Ming-zhi(Unit 61267 of PLA,Beijing 101114,China;Joint Operations College,National Defence University,Beijing 100091,China)

机构地区：[1]中国人民解放军61267部队,北京101114 [2]国防大学联合作战学院,北京100091

出　　处：《计算机仿真》2024年第10期16-21,共6页Computer Simulation

摘　　要：为辅助指挥人员进行决策,研究了大规模联合作战场景下的目标分群问题,提出了一种基于进化多目标优化的目标分群算法。采用了连续编码和解码方式,充分地利用了作战体系网络的结构特征,赋予分群问题一个较好的初始解;将反比例关联和比例缩减作为两个目标函数,把目标分群问题建模为一个多目标优化问题;在第二代非支配排序遗传算法框架的基础上结合差分进化算法设计了目标分群问题的优化算法。实验结果表明,上述算法能够有效地对战场上的目标进行分群。In order to assist commanders in making decisions,this paper makes researches on the grouping of operational targets in large-scale joint operations,an evolutionary multi-objective optimization algorithm for partioning operational targets into groups is proposed.By using continuous encoding and decoding,the structural characteristics of the operational system network are fully utilized and giving a better initial solution to the grouping problem.The task of the operational target grouping is modeled as a multi-objective optimization problem with the inverse ratio association and ratio cut are selected as the objective function.The optimization algorithm for operational target grouping is built upon the NSGA-Ⅱ with differential evolutionary operators.The experiment results show that the algorithm can effectively partitioning operational targets into groups.

关 键 词：目标分群 作战体系网络 多目标优化 连续编码 遗传算法 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术] E919[自动化与计算机技术—计算机科学与技术]