Tipping Dynamics and Critical Recovery in Interdependent UAV Swarm Networks

The rapid deployment of unmanned aerial vehicle (UAV) swarms across the low-altitude economy raises critical concerns regarding their operational safety and resilience, given complex dynamic environments and multi-dimensional risk couplings. Current research remains limited in adequate characterization of multi-layer interactions, lacks mission-priority-aware recovery mechanisms, and neglects critical recovery phases in evolutionary analysis. To address these gaps, this paper proposes a safety assessment framework that integrates dual-layer coupled network modeling with an improved SIRD (Susceptible-Infected-Recovered-Deleted) propagation model. Specifically, a "Structure-Mission"dual-layer network is constructed to capture both dynamic UAV interactions and mission collaboration logic, while the improved SIRD model incorporates mission-priority-differentiated state transition rules to simulate fault propagation and system recovery. Numerical simulations reveal that the system's evolution, from tipping dynamics to critical recovery, is governed by the recovery rate, network scale, and high-priority node recovery probability. This study deciphers the intrinsic connection between failure initiation and recovery patterns, providing a theoretical insight for proactive safety management and strategic optimization of UAV swarm operations in complex low-altitude environments.