Research On Quality Assessment For ADS-B Based Moitoring Data

As a core component to constitute future monitoring systems,it is necessary to accurately recognize and assess the abnormal condition and bug of the automatic dependent surveillancebroadcast(ADS-B)system.ADS-B data not only can be used to directly characterize the system’s operation performance,but also can be recognized as the theoretical support of various studies in the aerospace field.Therefore,it is of significant importance to assess the quality of monitoring data from the ADS-B based system.This paper concentrates on the quality assessment for ADS-B monitoring data and conducts the following studies.First,ADS-B data were pre-processed.based on C# programming language,the raw data totaling 71004319 items were resolved.And the interactive multi-model Kalman filtered trajectory was adopted as the reference trajectory by comparing and analyzing the simulation results.Second,from a distinct perspective,the inherent quality,operation quality and characterization quality were selected as the assessment criteria of data quality.Thus a versatile quality assessment system of ADS-B monitoring data based on Common Data Quality Assessment System was proposed and realized for the first time,to the best of our knowledge.Then,the quality assessment model was proposed and constructed on the basis of the assessment system.The data quality was ranked as 5 levels including excellent,good,fair,poor and very poor.Finally,on the basis of these resolved ADS-B data,each criterion was independently evaluated and then allocated through the training set.Subsequently,the random forest model was optimized by adjusting the hyper-parameters.The obtained result shows that as high accuracy as 96.48% was realized.Moreover,82.50% of the selected test data set satisfy the requirements of air monitoring systems.The achievements demonstrate that the proposed assessment scheme and random forest model is an effective way to assess the quality of ADS-B monitoring data,which provides a new route and theoretical foundation for quality assessment of ADS-B monitoring data.