Flight Data Analysis And Simulation Study Based On Data Mining

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.