Resumen
A careful arrival and departure sequencing of aircraft can reduce the inter-arrival/departure time, thereby opening up opportunities for new landing and/or take-off slots, which may increase the runway throughput. This sequence when serviced with a suitable runway configuration may result in an optimal aircraft sequence with a runway configuration that can process the maximum number of aircraft within a given time interval. In this paper, we propose a Cooperative Co-evolutionary Genetic Algorithm (CCoGA) to find the combined solution of a best-fit sequence with a feasible runway configuration for a given traffic demand at an airport. The aircraft sequence and the runway configuration are modelled as individual species, which can cooperatively interact with each other. Therefore, we computationally evolve the best possible combination of aircraft sequence (arrival and departure) and the feasible runway configuration. The proposed CCoGA algorithm is evaluated for Chicago O?Hare International Airport runway layout and resulting configurations. Arrival and departure traffic demand is modelled through a Poisson distribution. Two different arrival/departure sequencing methods, i.e., constraint position shifting with one, two and N-position shifting and first come first serve, are modelled. Runway configuration and traffic sequence (arrivals and departure) are modelled as two species, which are evolved co-operatively, through the CCoGA algorithm, to achieve the optimal traffic sequencing with a feasible runway configuration. Time-space diagrams are presented for the best-evolved population of arrival-departure sequence and runway configuration to illustrate the possibility of using available departure slots between arrivals to maximize capacity. Arrival-departure capacity envelopes are then presented to illustrate the trade-off between the arrivals and departures, given a runway configuration for each sequencing method. Results demonstrate the high mutual dependence between arrival-departure sequence and the runway configuration, as well as its effect on overall runway capacity. The results also demonstrate the viability of using evolutionary computation-based methods for modelling and evaluating complex problems in the air transport domain.