| To address the issues of low sampling efficiency and high sampling randomness in UAV path planning within complex 3D environments using the traditional Rapidly-exploring Random Tree (RRT) algorithm, this paper proposes an optimized method that integrates an improved Sparrow Search Algorithm (SSA) with RRT.First, during the RRT phase, a dynamic goal-biased sampling strategy is introduced. This strategy adaptively adjusts the probability of selecting the goal point to balance exploration and exploitation capabilities. Concurrently, an adaptive step-size mechanism, based on the distance to obstacles, is incorporated to enhance search efficiency.Subsequently, a chaotic initialization strategy using the Logistic map is employed to generate the initial population, thereby increasing population diversity. Furthermore, the Sigmoid function is utilized to dynamically adjust the weight coefficients for path length, smoothness, and safety, which strengthens the global exploration ability.Finally, a dynamic multi-objective optimization function is applied to optimize the metrics of path length, smoothness, and obstacle avoidance distance, thereby improving overall path quality.Simulation results demonstrate that the proposed fusion algorithm successfully reduces the number of nodes by 91.95%, decreases the average planning time by 60.87%, and shortens the path length by 13.25% compared to the traditional RRT algorithm. In complex 3D scenarios with multiple obstacles, the method efficiently generates safe and smooth optimized paths, exhibiting superior convergence speed and path quality over conventional algorithms. This provides a reliable solution for autonomous UAV navigation in complex environments. |
分享按钮