Analysis of the Trigger Factors of Radar Vectors Using Support Vector Machine for Arrival Traffic at Fukuoka Airport

Abstract

Fukuoka Airport is one of the most convenient airports in Japan. The airport has one runway and the congestion is normalized. It is keenly desired to be solved by all means. In this paper, the authors focused on the flow of arriving traffic at Fukuoka Airport, and analyzed the trigger factors of the radar vectors in the Fukuoka approach control area using CARATS Open Data and a Support Vector Machine which is a type of supervised machine learning. SHAP (SHapley Additive exPlanations) is used to visualize the analysis results of the trigger factors. The result showed that the simultaneous approach from the east and the west direction of the approach control area has the greatest effect on the start of the radar vector in the Fukuoka approach control area.