节点文献

基于数据挖掘的飞行数据分析及仿真研究

Flight Data Analysis and Simulation Study Based on Data Mining

【作者】 谢果

【导师】 魏麟;

【作者基本信息】 中国民用航空飞行学院 , 载运工具运用工程, 2012, 硕士

【摘要】 本论文主要来源于民航局项目——通用航空飞行数据采集与分析研究。本论文的主要工作分为两个大的方面:飞行数据分析和飞行3D仿真平台的建立。目前,飞行数据越来越被人们所重视,传统上飞行数据的分析方法较多,大多建立在飞机系统模型下,本文直接针对飞行数据,绕开飞机系统模型,运用数据挖掘技术的中聚类算法对飞行数据进行聚类分析,寻找出飞行数据之间内在的联系,并建立了聚类分析系统。该系统能够很好的把正常数据和异常数据区分开,对以后的日常机务维修和飞行安全警示都能起到一个较好的参考作用。同时,由于飞行数据的特殊性,较多数据包括相同属性。例如INS和GPS数据属性相同。本文提出了一种基于卡尔曼滤波的数据关联算法在飞行数据预处理上的运用,通过仿真实验证明该算法能有效的降低飞行数据的噪声,相对传统算法它具有收敛速度快的特点。同时它又为本文的3D仿真平台中的数据预处理奠定了基础。本文的最后一章在以上工作的前提下,利用VEGA平台建立了一套能够提供多视角的3D飞行数据仿真系统,并运用真实飞行数据对该系统进行了测试,该系统能较好实现飞行仿真播放控功能,飞行主姿态显示功能,飞行航迹显示功能,飞行横滚显示功能,飞行数据实时显示功能。

【Abstract】 This article originated from the Civil Aviation Authority Project–Research on generalaviation flight data acquisition and analysis.The main work of this article had been dividedinto two major aspects. One was that flight data had been analyzed, the other was that 3Dsimulation platform had been established.More and more attention has been paid to the flight data in nowdays. Traditionally, flightdata analysis was based on the flight system model of aircraft. Flight system model of aircraftwas not used in the article. Clustering Algorithm of Data Mining technology had been used inanalysis of flight data to find out the internal relationship of the data, and then ClusteringAnalysis System was established.Normal data and abnormal data could be well separated byusing the system, meanwhile it was a good reference for daily maintenance repair and flightsafety warning.Because of particularity of flight data,It includes the same attribute data. For example,the data in INS and GPS was the same.On this premise that flight data includes the sameattribute data, data association algorithm based on Kalmar filter had been proposed, whichcould be used in flight data preprocessing. According to the simulation experiment, thealgorithm could not only reduce the noise of flight data, but also had faster convergence ratethan that of the traditional algorithm. At the same time it laid the foundation for flight datapreprocessing of the 3D simulation.The last chapter of the paper based on the first few chapters introduces the developmentprocess. The 3D flight data simulation system was established, which could provide multipleangle observation.Real flight data was used to test this system. The system can realize theflight simulation playback control function, primary flight gesture display function, flight pathdisplay function, flight rolling display function and real time flight data display function.

【关键词】 VEGA数据挖掘聚类算法卡尔曼滤波
【Key words】 VEGAdata miningclustering algorithmKalmar filter
  • 【分类号】TP311.13;V247.5
  • 【被引频次】3
  • 【下载频次】357
  • 攻读期成果
节点文献中: 

本文链接的文献网络图示:

本文的引文网络