Abstract: The density-based spatial clustering of applications with noise-consensus-based bundle algorithm (DBSCAN-CBBA) is proposed for multiple unmanned aerial vehicles collaborative task allocation in static and dynamic rescue scenarios. First, considering the uncertainties in task execution and constraints such as payload limitations of unmanned aerial vehicles, a more realistic multi-task allocation model was established. Second, the task bundle construction structure of the consensus-based bundle algorithm was optimized to improve algorithm efficiency and the ability to find optimal solutions. In the initial phase, candidate task sets were constructed using the density-based clustering algorithm, and non-candidate task sets were constructed randomly; in the second stage, conflicts caused by independent task bundle construction among unmanned aerial vehicles were resolved through communication. Finally, the proposed algorithm was applied to both static and real-time dynamic task allocation scenarios. Simulation experiment results demonstrate that the proposed algorithm can efficiently and quickly find reasonable task allocation solutions.