A comprehensive evaluation method of ADS-B data quality based on clustering and AdaBoost

Traditional methods for ADS-B data quality assessment cannot obtain the quality level objectively and accurately. For better application of ADS-B data, an ADS-B data quality evaluation index system is constructed on the basis of analysis of ADS-B data quality requirements in industry applications, transmitting equipment performance, data security, etc. A new data quality evaluation method is proposed based on the ensemble learning Adaptive Boosting (AdaBoost) algorithm. In this method, the best quality grade category is determined by K-means clustering, data labels are determined by combining the Technique for Order Preference by Similarity to Idealsolution (TOPSIS), and the evaluation model is trained and optimized by the AdaBoost algorithm. The data of Tianjin Airport are use for case analysis. The experiment shows that it is the best scheme to divide ADS-B data quality into 5 grades, and the accuracy of the obtained data quality evaluation model is as high as 98.5%. This verifies that the method proposed can effectively avoid the influence of subjective factors and obtain the optimal quality grade classification, improving the stability and accuracy of evaluation results.