中国民航大学学报 ›› 2013, Vol. 31 ›› Issue (3): 14-18.

• 民用航空 • 上一篇    下一篇

基于改进的模糊C-Means航迹聚类方法研究

王超,王明明,王飞   

  1. (中国民航大学空中交通管理学院,天津300300)
  • 收稿日期:2012-07-04 修回日期:2012-08-28 出版日期:2013-06-23 发布日期:2013-10-08
  • 作者简介:王超(1971—),男,天津人,副教授,博士,研究方向为交通信息工程与控制,交通运输规划与管理.
  • 基金资助:

    国家科技支撑计划项目(2011BAH24B08);中国民航大学研究生科技创新基金(YJSCX12-03);中央高校基本科研业务费专项(ZXB2011A002)

Trajectory clustering method research based on improved Fuzzy C-Means

WANG Chao,WANG Ming-ming,WANG Fei   

  1. (College of Air Traffic Management,CAUC,Tianjin 300300,China)
  • Received:2012-07-04 Revised:2012-08-28 Online:2013-06-23 Published:2013-10-08

摘要:

为指导飞行程序的改善和发现管制员的指挥模式,在分析历史飞行航迹特征基础上,应用最小描绘长度(MDL)原理对航迹特征点进行划分,运用融合了遗传算法和模拟退火算法的改进的模糊C-Means 算法对特征点进行聚类,通过最长公共子序列(LCS)算法得到航迹相似性矩阵,利用矩阵得到航迹簇,最后形成中心航迹,算例仿真验证了新算法的有效性。

关键词: 航迹聚类, 遗传模拟退火算法, 模糊C-Means, 最长公共子序列

Abstract:

In order to guide the improvement of flight procedures and find the command mode of the controller,this paper firstly analyzes the historical flight path,and categorize the characters by the minimum description length (MDL) theory,then clusters the characters using of the improved Fuzzy C-Means (FCM) method combined with the GA and SA,and receives the trajectory similarity matrix by the longest common subsequence (LCS),lastly generates the mean routs through the trajectory clusters. The simulation tests show that this new methodis quite effective.

Key words: trajectory clustering, GASA algorithm, Fuzzy C-Means, longest common subsequence

中图分类号: