Presentation Information
[B-2-21]Curve Fitting of Departure Throughput of the Runway Using Machine Learning
〇Izumi Yamada1, Megumi Oka1 (1. Electronic Navigation Research Institute)
Keywords:
air traffic management,departure airport,machine learning,weather conditions,queuing phenomena
Efficient air traffic flow management requires that traffic demand at departure airports, which often constitute bottlenecks in the traffic flow, be managed in accordance with the actual number of departures that can be accommodated per unit time (departure throughput). Therefore, it is essential to develop a quantitative model that describes variations in departure throughput under different weather conditions. In this study, using Narita International Airport as a case study, we estimated departure throughput from aircraft takeoff/landing and block-out/block-in timestamp data, and investigated its relationship with METAR (Meteorological Aerodrome Report) and meteorological observatory data by applying machine learning–based fitting methods.