Abstract: To address the problem that conventional path planning methods tend to become trapped in local optima when the UAV performs tasks in dense-obstacle environments and multi-threat scenarios, resulting in insufficient trajectory safety and large fluctuations in mission cost, a UAV path planning method integrating multiple strategies is proposed. During the path generation stage, a Sobol sequence-based initialization mechanism is introduced to improve the distribution uniformity of candidate trajectories in three-dimensional space, thereby enhancing the UAV＇s global awareness of complex environments; during the flight decision process, an adaptive adjustment mechanism is constructed to enable the UAV to dynamically adjust its search behavior according to environmental changes and task progress, thereby improving its stability and adaptability under multi-threat interference. Meanwhile, drawing on the concept of cooperative search in swarm intelligence, the coordination between global exploration and local optimization during path planning is strengthened, enabling the UAV both to avoid large-scale risk areas and to finely adjust its flight trajectory in locally complex terrain. Furthermore, to address the tendency of UAV trajectories to cluster and exhibit insufficient flexibility in complex spaces, a spatially distributed elite-leading strategy is designed to guide the generation of diverse candidate trajectories, effectively improving path safety redundancy and interference resistance while avoiding dependence on a single path. Finally, experimental validation is conducted using typical test functions and a three-dimensional UAV flight simulation environment. The results show that the method can generate safer, more stable, and higher-quality flight paths in complex scenarios, significantly improving the UAV＇s mission execution capability and robustness.