A Direction Guided Dung Beetle Optimizer for Cooperative Multi‐ UAV Path Planning in Dynamic Environments

ABSTRACT Cooperative multi‐UAV path planning in dynamic environments faces severe challenges from high‐dimensional state spaces, strongly coupled spatiotemporal constraints, and the difficulty of simultaneously ensuring inter‐UAV collision avoidance and dynamic obstacle avoidance. To address these challenges, this paper proposes the direction guided dung beetle optimizer (DGDBO). Within a unified multi‐level directional regulation framework, an adaptive centripetal–centrifugal search strategy, a triangular deflection walk, and a quadratic interpolation‐based refinement operator are integrated to systematically coordinate exploration, local optima escape, and fine‐grained exploitation across different optimization stages. Furthermore, a continuous collision detection model based on spatiotemporal line segments unifies inter‐UAV and dynamic obstacle avoidance within a single analytical framework, ensuring flight safety while maintaining solution space connectivity. Evaluations on the CEC2017 benchmark suite show that DGDBO ranks first overall via the Friedman test, achieving optimal results on 25 of 29 functions with an average ranking of 1.172, outperforming the second‐best algorithm by approximately 1.97 in average ranking. In four dynamic multi‐UAV simulation scenarios of increasing complexity, DGDBO achieves the lowest mean path cost and smallest standard deviation among all six compared algorithms. Compared with the other competitor algorithms, DGDBO reduces mean path cost by an average of 16.4% and improves convergence stability by an average of 64%, with all performance advantages statistically confirmed by the Wilcoxon rank‐sum test. These results demonstrate that DGDBO provides a robust and efficient optimization solution for cooperative multi‐UAV path planning under complex dynamic constraints.