Airport utility stochastic optimization models for air traffic flow management

摘要

The complexity of air traffic flow management has its groundwork at an airport and increases with the number of daily aircraft departures and arrivals. To adequately contribute toward an accelerated air traffic flow management (ATfM), multivariate statistical models were developed based on airport utility. The utility functions were derived from daily probabilities of airport delay and inefficiencies computed using parameterized statistical models. The estimates were based on logistic and stochastic frontier models to derive distribution functions from which daily airport utilities were estimated. Data for testing and model simulations are daily aggregates spanning a five year period, collected from Entebbe International Airport. The utility models show that there was a 2 percent difference between daily aircraft operations at departures (92 percent) and at arrivals (94 percent). These findings confirm the likelihood that events leading to departures are more rigid compared to those observed at aircraft arrivals. Simulation results further confirmed that lowering delays at departure and arrival would result into higher airport utility. Airport utility was found to decrease consistently with an increase in the air-to-ground cost ratios. Airport utility analyses were most stable at a delay threshold of 60 percent and an air-to-ground cost ratio of 1.6 for both departures and arrivals. Therefore, for better outcomes of airport utility studies, this study recommends different treatments between departure and arrival analyses. The models developed are flexible and easily replicable with little adjustments to reflect airport specific characteristics. (C) 2014 Elsevier B.V. All rights reserved.