航空器轨迹变点检测与判识技术研究

发布时间：2018-03-30 08:38

  本文选题：航空排放　切入点：排放监测　出处：《中国民航大学学报》2017年04期

【摘要】：航空器飞行阶段的有效划分是支撑航空排放主动监测的重要技术手段,采用变点检测与判识的方法可有效地实现对航空器飞行阶段的划分。本研究根据多变量T~2统计量的特点,在数据整体参数一致性假设下,提出了一种采用顺序双滑窗的航空器轨迹变点检测与判识的方法。仿真与实验结果表明,该方法对窗口长度较为敏感,在适宜的窗口长度下可有效地解决轨迹的变点检测问题。根据航空器轨迹的特点,选择适当的窗口长度前提下,本方法是一种有效的变点检测与判识方法。
[Abstract]:The effective division of aircraft flight stage is an important technical means to support the active monitoring of aviation emission. The method of change point detection and recognition can effectively realize the division of aircraft flight stage. This study is based on the characteristics of multivariable TK2 statistics. Under the assumption of the consistency of the whole data parameters, a method for detecting and identifying the change points of aircraft trajectory using sequential double sliding windows is proposed. The simulation and experimental results show that the method is more sensitive to the length of the window. The change point detection problem can be effectively solved under the appropriate window length. According to the characteristics of aircraft trajectory and the premise of selecting appropriate window length, this method is an effective variable point detection and recognition method.
【作者单位】： 中国民航大学天津市智能信号与图像处理重点实验室;中国民航大学中欧航空工程师学院;
【基金】：中央高校基本科研业务费专项(3122017111)
【分类号】：V35;X738
，

本文编号：1685148