A High-Precision Cooperative Localization Method for UAVs Based on Multi-Condition Constraints

Global Navigation Satellite Systems (GNSSs) often suffer from significant localization errors in signal-denied environments. Furthermore, the accuracy of multi-UAV cooperative localization is highly sensitive to the relative geometric configuration of the swarm. To address these challenges, this paper proposed a novel high-precision and robust cooperative localization method for UAVs. The proposed method comprised two key modules. First, based on the principle of minimizing the Geometric Dilution of Precision, we optimized both the quantity and geometric configuration of the UAV swarm to identify the top three optimal aerial formations. Second, we introduced Ground-Assisted Reference Stations or Unmanned Ground Vehicles to establish an air–ground cooperative localization system. By leveraging Time Difference of Arrival constraints, this system significantly enhanced localization accuracy and robustness. From this analysis, two optimal hybrid configurations were selected. Experimental results showed that while purely air-based geometric optimization enhanced horizontal coverage, it failed to effectively suppress Z-axis errors due to inadequate vertical baselines, with deviations consistently oscillating between 3.0 m and 5.0 m. Conversely, the introduction of edge-deployed ground reference stations reduced the Position Dilution of Precision to a remarkably low level of 0.75, effectively suppressing error divergence. This demonstrated that the proposed air–ground cooperative scheme outperformed traditional pure air-based swarm approaches in localization performance. These findings hold significant theoretical and practical value.