Analyzing the Effect of Traffic Scenario Properties on Conflict Count Models

Decentralized en-route airspace concepts have been proposed by many studies to increase airspace safety and capacity. Most of these studies, including our own forays into this domain, have used fast-time simulation experiments to explore the benefits offered by decentralization. While simulations are indispensable during the initial design phase of any new airspace concept, the understanding gained using this approach can be difficult to generalize beyond the tested conditions. To address this issue, some researchers have presented analytical conflict count models to quantitatively analyze the effect of physical factors, such as traffic separation requirements, on the intrinsic safety of decentralized airspace concepts. However, the derivation of these models often make use of idealized assumptions regarding the behavior of traffic that do not always reflect realistic operations. To this end, this paper investigates the effect of these assumptions on the accuracy of the analytical conflict count models using targeted fast-time simulations of a direct-routing unstructured en-route airspace concept for a number of more realistic traffic patterns. The data collected from these simulations is also used to test so called ‘model adjustments’ that aim to relax the dependency of the models on the idealized traffic scenario assumptions. The results show that the assumptions do affect the accuracy of the analytical models, with some assumptions leading to a substantial under-estimation of conflicts. The results also show that the model adjustments increased accuracy for the more realistic scenarios to the levels previously found for the ideal traffic settings for all cases. Therefore, in addition to providing a physical understanding of the factors that affect airspace safety, the adjusted models can also be used as tools for practical airspace design applications. Keywords—Airspace safety; airspace design; conflict rate; conflict