A 0-1 MIP-based group-and-release strategy for solving the integrated runway scheduling and taxiway routing problem

A 0-1 MIP-based group-and-release strategy for solving the integrated runway scheduling and taxiway routing problem

Jitamitra Desai, S. Srivathsan, L. Li, W. Y. Lai, and C. Yu

Journal: Naval Research Logistics

With growing air traffic demand and the required airport infrastructure lagging behind by at least a decade, it has become imperative for air traffic controllers to efficiently squeeze the available capacity at an airport in order to minimize aircraft delays. It has been well documented that the two major bottlenecks affecting the smooth functioning of air traffic operations at an airport are runways and taxiways.

The key problem involving these resources includes the scheduling of flights on the runway, and the determination of the taxiway paths to be traversed by flights from their assigned gates to the runway.

The researchers address this problem by modeling an integrated runway scheduling and taxiway routing problem as a 0-1 mixed-integer program (MIP) in a free-path setting where any feasible taxiway route can potentially be assigned to a flight. As a direct application of this MIP model is not suitable for solving large-scale instances, they develop a three-step group-and-release strategy that first segregates the flights based on their allocated gates and associated ramps and then solves the MIP model for each ramp to determine the taxiway path for each flight. In the final step, the path for each flight is fixed, and a sequencing problem over all flights is solved to determine high quality, feasible solutions.

The performance of the proposed methodology is benchmarked against three algorithms, namely: (i) constraint-generation; (ii) sequential two-stage algorithm; and (iii) FCFS algorithm. Their numerical experiments, based on actual flight data from Changi airport (Singapore), show that, on average, the optimality gap as well as the computational time is considerably reduced for our strategy as compared to existing methods, thereby highlighting the efficacy of the proposed approach in solving realistic instances.

EASYSENDY.com