基于聚类分析的武器目标分配问题求解  

Solving Weapon-Target-Assignment Problem Based on Cluster Analysis

作　　者：靳萌[1] 陈立云[1] 

机构地区：[1]军械工程学院,河北石家庄050000

出　　处：《科技广场》2009年第7期15-18,共4页Science Mosaic

摘　　要：利用伪并行遗传算法与K-均值聚类算法结合解决武器目标分配问题,将目标威胁值的分配问题转化为目标的分类问题。在采用聚类分类算法的基础上,使用伪并行遗传算法对分类结果进行优化,克服了K-均值聚类算法分类局限性,提高了全局搜索能力,达到了局部收敛速度与全局收敛性能的统一。在已知目标威胁值的情况下,利用遗传算法完整解决了WTA问题。通过仿真程序实现,验证了算法的可行性,由此为作战仿真CGF技术中的武器目标分配问题的解决提供了方法。Make use of pseudo parallel genetic algorithm and K-meansclustering to combine a solution of weapon target assignment problem. Allotment of threaten value of the target can be transformed to the classification problem of targets. When the targets had been classified,the result of cluster analysis was optimized by pseudo parallel genetic algorithm. Overcame the limit of K-means clustering,improve the ability of macroscopical searching and reach the unification of speed of partial astringency and the capability of macroscopical astringency. Made use of genetic algorithm to solve WTA the problem is under the sistuation that have already known the threaten value of the targets. The feasibility of the algorithm have been validated through the implement of emluator. The algorithm provided a solveing methd of WTA of the CGF of the battle simulation.

关 键 词：Weapon-target-assignment遗传算法 聚类分析 

分 类 号：TP301.66[自动化与计算机技术—计算机系统结构]