低空立体交通跨模式协同与智能调度研究综述

As the advancement of low-altitude three-dimensional transportation systems accelerates, the collaboration among aerial, ground, and waterborne heterogeneous transportation platforms becomes increasingly critical. The seamless integration of intelligent scheduling and resource optimization is progressively establishing itself as an indispensable pillar for the development of smart cities and emergency response frameworks. A comprehensive review of the research frontiers in multi-modal collaborative scheduling is provided. Firstly, it comprehensively reviews the collaborative mechanisms of unmanned aerial vehicles, unmanned ground vehicles, and unmanned surface vehicles in multi-source information fusion, environmental perception, and adaptive decision-making. Secondly, this paper synthesizes typical scheduling methodologies in complex dynamic environments, addressing task decomposition, path planning, collaborative control, and system-level scheduling, with a particular focus on reinforcement learning, graph optimization, and evolutionary algorithms. Furthermore, it explores task allocation and communication strategies within centralized, distributed, and hybrid control architectures, and examines the role of sensor data, simulation data, and operational big data in optimization of scheduling processes. Finally, the critical challenges faced by current low-altitude three-dimensional transportation systems in resource allocation, security assurance, and cross-domain collaboration are identified, and insights into the future development trajectories of these systems, underpinned by big data-driven solutions and intelligent enhancement, are offered.