Centralized and distributed optimization of advanced air mobility strategic traffic management

Effective traffic management for advanced air mobility (AAM) operations in low-altitude urban airspace is crucial for safety and scalability. Our study aims to bridge a critical gap in AAM traffic management by minimizing travel delays in both centralized and distributed providers of services for urban air mobility (PSU) settings. Key contributions include methods to (1) sectorize urban airspace for effective AAM management, (2) centrally plan AAM routes considering limited capacities in corridors and vertiports, and (3) manage airspace in distributed PSU settings while considering traffic flow capacities and interactions among PSUs. Specifically, the research combines community detection algorithms with Voronoi diagrams to sectorize individual PSU airspace. Corridor route planning is performed with a custom-weighted Dijkstra’s algorithm. Centralized AAM traffic flow management adopts mixed-integer programming (MIP) to minimize overall network delay costs. Distributed PSU network management is formulated as bi-level optimization using cooperative game theory and MIP, where individual PSUs update their strategies based on game theory outcomes. The simulation environment features a randomized no-fly zone, population density maps, and vertiport capacities assigned to artificial cities. Three vehicle configurations with varying ranges and adjustable speeds (i.e., minimum to cruise speeds) are simulated under three service priorities in Monte Carlo simulations. AAM flight operations are evaluated by optimization cost and runtime. This research provides a technical framework and insights into the comparison of centralized and distributed AAM network managements. The paper will facilitate informed decision-making in the development and implementation of AAM traffic management strategies.