文章摘要
张鹏,黄长强,魏政磊,周欢,王永乾.基于L Kshape HACA的空战态势分割聚类[J].空军工程大学学报:自然科学版,2021,22(3):15-22
基于L Kshape HACA的空战态势分割聚类
Air Combat Situation Segmentation Clustering Based on L Kshape HACA
  
DOI:
中文关键词: 无人自主空战  层次聚类分析  多元态势提取  拉普拉斯中心性方法
英文关键词: close air combat  hierarchical cluster analysis  multivariate situation extraction  laplace centrality method
基金项目:陕西省自然科学基金(2020JQ-481)
作者单位
张鹏,黄长强,魏政磊,周欢,王永乾 1.空军工程大学航空工程学院 西安 710038 2.95596部队河南商丘476000 
摘要点击次数: 10
全文下载次数: 22
中文摘要:
      在近距空战中,实时获取可靠的空战态势信息对于决策指引是非常重要的。针对近距空战态势变化剧烈以及评估参数多维耦合的问题,提出了一种基于L-Kshape HACA的多元空战态势分割聚类方法。以分层时序聚类分析为框架,利用拉普拉斯中心性方法确定聚类数目,同时采用Kshape对多元时间序列进行聚类分析,解决了多维参数下的态势信息提取问题。利用12组近距空战数据进行测试,并与14种聚类算法进行比较,结果表明L-Kshape-HACA在聚类中心确定和态势分割准确性上与实际空战态势更加符合。
英文摘要:
      In close air combat, it is very important to get a piece of reliable air combat situation information in time for decision making guidance. Aimed at the problems that there are drastic changes in the situation of close air combat and multi dimensional coupling of evaluation parameters, a multi element air combat situation segmentation clustering method based on L Kshape HACA is proposed. Taking the hierarchical time series cluster analysis as a framework, the number of clusters is determined by using the Laplace centrality method, and the cluster analysis on multivariate time series is made by using Kshape, solving the problem of situation information extraction under multi dimensional parameters. The test is performed by using the 12 sets of close air combat data, and the 14 clustering algorithms are compared. The results show that L Kshape HACA is more consistent with the actual air combat situation in the accuracy of cluster center determination and situation segmentation.
查看全文   查看/发表评论  下载PDF阅读器
关闭