Optimal Dynamic Scheduling of Aircraft Arrivals at Congested Airports

Abstract

An optimal airport arrival scheduling algorithm, which works within a hierarchical scheduling structure, is presented. This structure consists of schedulers at multiple points along the arrival route. Schedulers are linked through acceptance rate constraints, which are passed up from downstream metering points. The innovation in this algorithm is that these constraints are computed by using an Eulerian model-based optimization scheme. In this scheme an aggregate airspace model is derived online. Optimization based on predictions of this model is used to compute the optimal acceptance rates at all metering points, so that the airport throughput is maximized while keeping sector counts within limits. This rate computation removes inefficiencies introduced in the schedule through ad hoc acceptance rate computations. The scheduling process at every metering point uses its optimal acceptance rate as a constraint and computes optimal arrival sequences by using a combinatorial search algorithm. We test this algorithm in a dynamic air-traffic environment, which can be customized to emulate different arrival scenarios. Results from Monte Carlo simulations show improved performance over first-come, first-served ordering for different values of key parameters such as arrival aircraft mix, average arrival rate, maximum position-shift constraint, and level of uncertainty in estimated time of arrival prediction.