A Multi‐Strategy Enhanced Dung Beetle Optimization Approach for Cooperative Path Planning in Multi‐ UAVs Systems

Cooperative path planning for multi‐unmanned aerial vehicles (UAVs) is a challenging and computationally intensive optimization task, which requires the coordinated generation of feasible and high‐quality trajectories under complex environmental constraints. The dung beetle optimization algorithm (DBO) provides a viable approach to addressing this problem; however, its global search capability is relatively limited and it is prone to being trapped in local optima. To address these challenges, this paper proposes a multi‐strategy enhanced DBO (MSDBO) approach for cooperative path planning in multi‐UAVs systems. First, a mathematical model incorporating multiple constraints, including flight distance, threat avoidance, and spatial cooperation, is established to transform the complex multi‐UAVs cooperative path planning problem into a multi‐constrained optimization problem. Subsequently, MSDBO is developed by integrating four innovative strategies: (1) Latin hypercube sampling is employed to enhance population diversity during initialization; (2) bidirectional guided search strategy is introduced into the rolling phase to strengthen global exploration and improve the ability to escape local optima; (3) adaptive spiral search strategy is incorporated into the foraging phase to better balance global exploration and local convergence; and (4) longitudinal crossover–mutation strategy is applied in the later stage, performing cross‐dimensional perturbations among different dimensions of elite individuals to alleviate high‐dimensional variable coupling and maintain population diversity. Through the synergistic interaction of these strategies, DBO achieves a more effective dynamic balance between exploration and exploitation, thereby significantly improving its overall optimization performance. The performance of MSDBO is evaluated using the CEC2017 and CEC2022 benchmark functions and compared against nine competing algorithms. Experimental results indicate that MSDBO achieves superior solution accuracy, convergence speed, and stability in numerical optimization tasks. Further experiments conducted in four distinct multi‐UAVs path planning environments demonstrate that MSDBO obtains lower total costs and generates higher‐quality cooperative trajectories, reflecting strong robustness and practical engineering applicability.