A Multiobjective Optimization Approach for Air Traffic Flow Management for Airspace Safety Enhancement

Abstract

This work aims to enhance the safety of air trafc in a procedural airspace by air trafc ow management (ATFM) without compromizing air trafc demand. In civil aviation, collision risk is an important indicator for air trafc safety assessment. In this work, we propose a generic ATFM framework based on multiobjective optimization for reducing lateral collision risk of the trafc in a given procedural airspace. The proposed framework aims to optimize the ight level assignment for a given set of ight plans such that the lateral collision risk can be reduced. To achieve this goal, we formulate an optimization problem containing two partially conicting objective functions. We then adopt three well-known evolutionary algorithms, i.e., MODPSO, NSGA-II, and MOEA/D to solve the optimization problem. We specially design some of the operators of those algorithms to make them suitable for the optimization problem. We merge the solutions yielded by those three algorithms and lter out the nal Pareto solutions. We carry out a case study on the procedural airspace of Singapore ight information region (FIR) with respect to twelve daily trafc data selected from the real trafc data for December 2019. Experiment results demonstrates that the lateral occupancy which is the key contributor to lateral risk can be reduced by 10.65% to 93.05% at a strategic planning level. This research contribute to strategic ight planning by assigning ight levels that may reduce the risk of collision in procedural airspace.