摘要
针对敌对伪装目标潜入我重点海域进行敌对活动时,难以利用民用手段或雷达特性区分识别的问题。通过分布式并行处理积累的雷达数据,总结各类目标活动规律,能够分析对比发现伪装目标。针对雷达数据的海量特性,使用分布式处理框架Hadoop处理原始雷达数据,生成轨迹知识库,并以知识库为基础利用异常轨迹分析算法判断轨迹是否异常。结果表明,该方法大约3天实现了全年雷达数据的异常轨迹分析,并能通过增加节点进一步提升分析速度。
It＇s difficult to use civil means or radar characteristics to distinguish Hostile target diving our key sea areas for hostile activities. Through the distributed parallel processing of radar data, summing up the rules of various types of target activities can analyze and find the camouflage targets. In view of the massive characteristics of radar data, the distributed processing framework Hadoop is used to process the raw radar data. The path knowledge bases are generated, which is used to determine whether the trajectory is abnormal with the abnormal path analysis algo- rithm. The results show that the proposed method can achieve the abnormal trajectory analysis of the whole year radar data within about three days, and the analysis speed will increase by increasing the node.
作者
孟凡君
曹伟
管志强
MENG Fanjun CAO Wei GUAN Zhiqiang(Human Resources Department, Nanjing Marine Radar Institute, Nanjing 210000, China)
出处
《电子科技》
2017年第1期41-45,共5页
Electronic Science and Technology
关键词
雷达数据
分布式存储
并行处理
异常轨迹
radar data
distributed storage
parallel process
abnormal path
