Skip to main content

Identification of Traffic Patterns and Selection of Representative Traffic Samples for the Assessment of ATM Performance Problems

Paper ID

SIDs-2022-089

Conference

SESAR Innovation Days

Year

2022

Theme

ATM Concepts and Operations

Project Name

SESAR 2020 ER4 project SIMBAD

Keywords:

ATM, clustering, k-means, Machine learning, traffic patterns

Authors

Raquel Sánchez, David Mocholí, Oliva García Cantú, Ricardo Herranz, Rubén Rodrígez, Faustino Tello and Adrián Fabio

DOI

Project Number

894241

Abstract

Despite being the only reliable way to assess the impact of future ATM solutions, the complexity of large-scale, bottom-up microsimulation models is often a barrier for their effective use to support decision-making. As a consequence, in many cases the simulations are limited to one or few particular days, usually selected based on expert judgement and/or simple rule-of-thumb criteria (e.g., simulate the day with the highest number of scheduled flights). This may not be representative of the impact of a given operational improvement under all possible traffic scenarios, especially considering the extreme complexity of the European airspace, with significantly different traffic flows on different days of the year in terms of traffic conditions. Hence, a realistic representation of traffic demand patterns is an essential condition for a comprehensive evaluation of new concepts, which may deliver very different performance gains depending on the level of traffic density and complexity. This paper proposes a methodology for the identification of traffic patterns and the selection of representative traffic samples (representative days) for the assessment of a specific ATM performance problem.