摘要
设计和实现了一种改进的蚂蚁聚类算法.基于海上空袭目标攻击方向划分问题,分析了传统的聚类算法解决此类问题的不足,提出了一种动态调整的空袭方向划分混合蚂蚁聚类算法.该算法能充分利用空中目标信息动态调整参数,以获取合理聚类数和加速算法收敛,对孤立数据处理的鲁棒性较强.用人工数据集和真实数据集进行实验.结果表明,该算法是一种高效率的聚类算法,提高了空袭方向划分的准确性和科学性.
A dynamic alignment hybrid ant-clustering and k-medoids （DAACM） algorithm is designed and realized. Based on aerial attack directions judgment problem, the deficiencies of traditional clustering algorithms are analyzed, DAACM algorithm for solving this problem is proposed. The novel antclustering algorithm can make full use of aerial target information, dynamically align parameters to gain the reasonable number of clusters and accelerate convergence. Also, when dealing with the isolated data, DAACM has good robustness. Some experiments have been made on real data sets and synthetic data sets. The results demonstrate that DAACM is an precision and reasonableness in solving aerial attack directions effective algorithm and can improve the judgment problem.
出处
《北京理工大学学报》
EI
CAS
CSCD
北大核心
2006年第6期504-507,511,共5页
Transactions of Beijing Institute of Technology
基金
国家部委预研项目(10504033)
关键词
蚂蚁聚类
动态调整
空袭方向划分
ant-clustering
dynamic alignment
aerial attack directions judgment
