Aircraft Conflict Resolution: A Benchmark Generator

TL;DR

This paper introduces a flexible benchmark generator for aircraft conflict resolution, enabling standardized testing of optimization methods across diverse, realistic scenarios to improve comparison and evaluation.

Contribution

It provides a novel benchmark generator that creates heterogeneous, realistic test instances for aircraft conflict resolution, facilitating fair comparison of optimization approaches.

Findings

Supports predefined, pseudo-random, and random scenarios

Allows testing under various difficulty levels

Enables comprehensive evaluation of optimization methods

Abstract

Aircraft conflict resolution is one of the major tasks of computer-aided air traffic management and represents a challenging optimization problem. Many models and methods have been proposed to assist trajectory regulation to avoid conflicts. However, the question of testing the different mathematical optimization approaches against each other is still open. Standard benchmarks include unrealistic scenarios in which all the flights move toward a common point or completely random generated instances. There is a lack of a common set of test instances that allows comparison of the available methods under a variety of heterogeneous and representative scenarios. We present a flight deconfliction benchmark generator that allows the user to choose between (i) different predefined scenario inspired by existing benchmarks in the literature; (ii) pseudo-random traffic meeting certain congestion measurements; (iii) and randomly generated traffic. The proposed setting can account for different levels of difficulty in the deconfliction of the aircraft and allows to explore and compare the real limitations of optimization approaches for aircraft conflict resolution.