An Information Entropy‐Based Adaptive Flower Pollination Algorithm for Unmanned Aerial Vehicle Path Planning

UAV path planning is a challenging optimization problem, in which feasible trajectories are required to simultaneously satisfy multiple constraints, including path length, flight altitude, energy consumption, and obstacle avoidance. Swarm intelligence optimization algorithms have been widely applied to solve such problems due to their simple structures and favorable optimization performance. However, these algorithms generally suffer from slow convergence, insufficient population diversity in the later search stage, and an imbalance between exploration and exploitation. To address these issues, an information entropy‐based adaptive flower pollination algorithm (AFPA) is proposed for UAV path planning. To balance the exploration and exploitation processes of the algorithm, an information entropy‐based adaptive switching strategy is developed to dynamically regulate the execution ratio between global and local pollination. In the global pollination stage, a spiral search mechanism is incorporated to enhance position updating, thereby improving population diversity and global exploration capability. In the local pollination stage, a critical‐dimension‐based update strategy is introduced to accelerate convergence and further improve solution quality. Experimental results on the CEC2017 benchmark suite demonstrate that AFPA outperforms the comparison algorithms in terms of convergence speed and optimization accuracy. Finally, a three‐dimensional UAV path planning environment is constructed based on the elevation data of Christmas Island. The results show that the proposed algorithm can generate safe, smooth, and high‐quality flight paths while satisfying multiple constraints, thereby demonstrating its effectiveness and application potential in complex real‐world environments.