基于双层遗传算法的静态防空反导杀伤网构建方法  

Static Air Defense and Antimissile Kill-Net Construction Method Based on Double-Layer Genetic Algorithm

作　　者：付昱龙 张海宾 郭相科[2] 戚玉涛 FU Yulong;ZHANG Haibin;GUO Xiangke;QI Yutao(School of Cyber Engineering,Xidian University,Xi’an 710126,China;Air Defense and Antimissile School,Air Force Engineering University,Xi’an 710051,China)

机构地区：[1]西安电子科技大学网络与信息安全学院,西安710126 [2]空军工程大学防空反导学院,西安710051

出　　处：《空军工程大学学报》2025年第1期59-66,共8页Journal of Air Force Engineering University

基　　金：国家自然科学基金(61772392)。

摘　　要：在防空反导作战中,构建高效的杀伤网是确保反导任务成功的关键环节,然而杀伤网构建问题的优化建模和求解是一个难题。从组合优化的角度出发,对静态的杀伤网构建问题进行优化建模并提出高效求解方案;针对杀伤网构建问题的特点,建立了混合整数规划模型,并使用一种基于双层规划的优化建模方案进行简化,通过任务分配和冲突消解的主从问题协同降低求解难度,随后设计一种基于双层遗传算法的求解框架进行优化实验。在4组不同规模的环境实验测试中,算法能够快速找到较优解,并且对较大规模问题能保持良好的求解能力以及良好的可解释性。为防空反导领域的静态的杀伤网自主智能构建提供思路,可作为动态杀伤网调整的研究基础。In air defense and antimissile operations,constructing an efficient kill-net is a key part to ensuring the success of air defense tasks,while it is tricky to model and to optimize such problems.This paper aims to model the construction of static kill-net from the perspective of combinatorial optimization and propose efficient optimization methods.Considering the characteristics of the kill-net construction problem,this paper establishes a mixed-integer programming model and uses a bilevel optimization modeling scheme for simplification.By the cooperation of the leader problem of tasks assignment and the follower problem of resolving conflicts,the optimization difficulty is reduced.Subsequently,a solution framework based on a bilevel genetic algorithm is designed.In experimental tests on 4 sets of different-scale environments,the algorithm is able to get a fine result quickly,with great interpretability and a good capabilities for solving larger-scale problem.This work provides insights for the autonomous intelligent construction of static kill-net in the field of air defense and could serve as the basis for research focusing on dynamic kill-net adjustment.

关 键 词：混合整数规划 杀伤网 遗传算法 防空反导 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程] E919[自动化与计算机技术—控制科学与工程]