Multiobjective Optimization Framework for Complex Interdependent Airspace Conflict Resolution

As diverse aircraft increasingly operate in shared airspace, resolving airspace conflicts under complex interdependent requirements has become a core challenge in air traffic management (ATM). Conventional methods struggle to achieve efficient high-quality solutions in high-dimensional multiobjective optimization (MOO) with multiple constraints and strong temporal correlations across airspace groups. This study introduces an optimized airspace conflict resolution method constructed using an improved nondominated sorting genetic algorithm III (NSGA-III) within a three-layer progressive framework. First, the conflict detection layer integrates interval coincidence detection with the separating axis theorem to achieve precise identification of 3D airspace attribute conflicts. Second, the conflict resolution layer establishes a constrained high-dimensional MOO model that minimizes airspace priority, types, quantities, and magnitudes of airspace adjustments. Finally, the airspace group optimization layer introduces neighboring airspace concepts and a dynamic airspace rejection strategy, which are implemented through improved NSGA-III to address strong inter-airspace constraints and domino effects. For NSGA-III, we propose an airspace-group-oriented encoding–decoding mechanism, a conflict-pair-based penalty function design, and a dynamic rejection strategy that jointly enhance convergence and solution diversity in high-dimensional objective spaces. Simulation experiments using real-world airspace demand data show that the proposed method achieves complete conflict resolution among retained demands (100%; mean rejection rate: 7.15%), with lower rejection costs, higher mean HV (1.14 versus 0.49, 0.51, and 0.96), and mean runtime reductions of 78.6%, 35.4%, and 48.5% relative to MOCOA, MORIME, and MOEA/D, respectively. Overall, the proposed method provides a robust and efficient paradigm for intelligent optimization of structured demand-based airspace systems.