An evolutionary computational framework for capacity-safety trade-off in an air transportation network

Abstract

Abstract Airspace safety and airport capacity are two key challenges to sustain the growth in Air Transportation. In this paper, we model the Air Transportation Network as two sub-networks of airspace and airports, such that the safety and capacity of the overall Air Transportation network emerge from the interaction between the two. We propose a safety-capacity trade-off approach, using a computational framework, where the two networks can inter- act and the trade-off between capacity and safety in an Air Transport Network can be established. The framework comprise of an evolutionary computation based air traffic scenario generation using a flow capacity estimation module (for capacity), Collision risk estimation module (for safety) and an air traffic simulation module (for evaluation). The proposed methodology to evolve air traffic scenarios such that it minimizes collision risk for given capacity estimation was tested on two different air transport network topologies (random and small-world) with the same number of airports. Experimental results indicate that though airspace collision risk increases almost linearly with the increasing flow (flow intensity) in the corresponding airport network, the critical flow depend on the underlying network configuration. It was also found that, in general, the capacity upper bound depends not only on the connectivity among airports and their individual performances but also the configuration of waypoints and mid-air interactions among conflicts. Results also show that airport network can accommodate more traffic in terms of capacity but the corresponding airspace network cannot accommodate the resulting traffic flow due to the bounds on collision risk.