Strategy Optimization of Low-Altitude Logistics Systems based on Belief Reliability

With the rapid growth of urban demand for instant delivery, coordinated UAV swarm delivery systems have emerged as a key technological pathway for fulfilling highefficiency logistics tasks. This study addresses the scheduling challenge under high-frequency order generation scenarios by proposing an order dispatching strategy based on hierarchical classification of remaining time and delivery distance. The strategy dynamically allocates limited UAV resources according to each order's remaining schedulable time and distance to the destination, aiming to improve the overall on-time completion rate. To ensure the accuracy of reliability measurement under small-sample data conditions, belief reliability theory is introduced, allowing for a more robust assessment of system performance. Within the developed simulation platform, the system continuously generates random delivery orders characterized by attributes such as weight, location, and generation time. A comparative analysis is conducted between the proposed optimized scheduling strategy and a conventional sequential dispatching approach. This framework captures the degradation of system reliability over time under different performance threshold settings. Simulation results demonstrate that the proposed strategy significantly slows down the decline in belief reliability under medium and high threshold conditions, thereby enhancing the stability and service quality of the UAVbased delivery system.