基于Fisher聚类算法的航班飞行高度偏差分析研究

目前计算机飞行计划(Computerized Flight Plan,CFP)中的计划飞行高度与实际飞行高度之间存在偏差.为了能识别出航路中易发生显著偏差的子航段,便于对其进行重点监控,本文提出一种基于Fisher有序样本聚类算法的航班飞行高度偏差分析方法,采用修正的Fisher有序样本聚类算法对航路进行分段,在分段的基础上,对每个子航段上的高度偏差平均水平及波动情况进行统计分析,作为高度偏差的评价指标.分析结果可帮助识别航路上易发生显著实际与计划飞行高度偏差的子航段,为签派人员制作飞行计划以及在机组人员进行航前准备提供决策依据,具有较好的实践意义.

A Study on Flight Altitude Discrepancy Base on the Fisher Ordinal Samples Cluster Method

There now exists a discrepancy between the practical flight altitude and the one calculated by CFP(Computerized Flight Plan),which is unfavorable to security as well as cost-saving of flights.Along air route,flight altitude discrepancy is variable between segments.According to this character,this paper studies the flight altitude discrepancy based on air route segmentation.The study applies the weighted Fisher ordinal samples cluster method to cluster around those air route points with similar discrepancy level,meanwhile keeping their original order in air route,consequently carrying out air route segmentation.The out come cantell on which air route segments the discrepancy is striking and needs to be given special at tention,which is of benefit to improve the rationality of flight plan(takeoff fuel) and helps pilots support takeoff preparation.