A Data-Driven Dynamic Modeling of Airport Runway Queuing System

The airport runway queue system is a complex dynamic system that continuously changes with the takeoffs and landings of aircraft. Enhancing the precision of simulation modeling for this system is crucial for accurately evaluating airport operational efficiency at both strategic and pre-tactical levels. However, the existing modeling methods based on queuing theory lack refinement in capturing the uncertainties of flight demand and airport service capability. This leads to significant discrepancies between simulation results and actual operations. Therefore, this paper proposes a data-driven approach to establishing the airport runway queue system and employs Monte Carlo simulation to model the dynamic queuing process of arriving and departing flights. The clustering method analyzes historical operational data to uncover demand and service capability patterns. To address global demand fluctuations, the flight demand statistics are derived from historical data and combined with scheduled flight data to construct probability distributions for each demand pattern. A hidden Markov model represents the time-varying transition characteristics of service capability for time-dependent service capability. Using Nanjing Lukou International Airport in China as a case study, the results show that the estimation errors for demand and service capability are below 5% and the simulated flight delay levels closely match the actual delay levels.