A method of cleaning ads-b flight track data based on fuzzy clustering

Abstract

The invention relates to a method for cleaning ADS‑B flight track data based on fuzzy clustering. The steps are as follows: data deduplication; feature field screening; abnormal data detection and analysis - fuzzy clustering method; time stamp correction; target field selection, fuzzy clustering method to detect abnormal track points and time stamp correction. The fuzzy clustering method can be used to quantitatively determine the data relationship, remove abnormal data, correct the flight trajectory, and then correct the time stamp through trajectory calibration to make the flight trajectory conform to the performance law. This method starts from the real flight situation, and has stricter requirements on data processing, so that the cleaning accuracy is higher, the exclusion effect is better, and the cleaning method is more accurate.

Description

technical field [0001] The present invention relates to a kind of ADS-B abnormal data detection method, particularly a kind of ADS-B flight track data cleaning method based on fuzzy clustering, used for cleaning the flight track data accepted by ADS-B ground station equipment, so that for Airlines analyze flight operations, air traffic control units evaluate airspace operation quality, and airport flow monitoring provides more accurate flight trajectory data, making data analysis and evaluation more accurate. Background technique [0002] With the rapid development of my country's economy and civil aviation industry, the number of aircraft is increasing year by year, and the burden on airspace is becoming more and more serious. The air traffic control department also lacks effective monitoring and communication means for aircraft within its jurisdiction, which seriously affects and restricts flight safety and operational efficiency. In this context, the ADS-B system based o...

AI Technical Summary

Problems solved by technology

[0007] The existing cleaning methods for ADS-B flight trajectory data usually use DBSCAN density clustering and other commonly used clustering methods in the detection of abnormal track points. For the DBSCAN density clustering method, if the density of the samples is not If the distance difference is very different, then the reliability of clustering will be relatively poor, that is to say, the DBSCAN density clustering algorithm is not a completely stable algorithm

[0008] Secondly, in the existing ADS-B data cleaning methods, the elimination methods for points that conflict with the longitude and latitude position and the timestamp sequence before the timestamp correction are mostly simple restrictions and judgments, so that the fault tolerance of the algorithm will increase, and the exclusion effect will be better. Difference

Images