摘要
飞行安全是飞行运输中的一个重要课题,对飞行纪录数据进行分析是判明飞行事故原因的重要依据,所以对飞行记录数据进行数据挖掘则成为一种有效的飞行监控技术。该文针对具有高维特征和大样本数据集的飞行纪录的飞行纪录数据学习问题,提出了一种新型的判别异常飞行事件的特征提取方法,并通过实验表明该方法对某一机型的飞行纪录数据取得了良好的实验结果。
Flight safety is an eternal subject of civil aviation transport worldwide. Quick Access Recorder （QAR） is an important gist for distinguishing the reason of flight incidents, so the intelligent analysis for it is a very significant task. For high-dimensional feature of the QAR data, and the large sample learning problem. It＇ s hard to learn effective and efficient. This paper proposes ball vector machine feature extraction method for QAR data suitable for discriminating abnormal flight incident. Finally, the experiment shows the algorithm have a good performance in the real dataset from a type of aircraft.
出处
《信息化研究》
2011年第2期68-71,共4页
INFORMATIZATION RESEARCH
关键词
空中交通
快速存取纪录器
特征提取
最小闭包球
air traffic
quick access recorder
ball vector machine
feature extraction
minimum enclosing ball
