基于BI-APF-RRT算法的无人机路径规划
UAV Path Planning Based on BI-APF-RRT Algorithm

针对双向RRT算法中存在较大搜索范围随机性问题，论文提出一种基于目标偏置的BI-APF-RRT (Bidirectional Artificial Potential Field Method RRT)改进算法的无人机飞行路径规划策略。首先，将目标偏置策略用于随机采样点的产生和双树扩张方向的指导，并通过双向RRT算法建立两棵相互交替的随机搜索树完成搜索，实现了算法收敛速率的提升。其次，将改进的人工势场融入双向生长树，进一步较大程度减少了搜索迭代次数。同时，在平滑路径的过程中，通过采用3次B样条插值算法实现了轨迹路径的优化。最后，通过仿真实验表明：与一些已有算法相比，基于目标偏置的策略改进BI-APF-RRT算法能够有效减少迭代次数，提升收敛速度，同时改善了新节点生成的方向，有效降低了路径成本，较好地解决了双向RRT算法中存在较大搜索范围随机性问题。
To address the issue of significant search space randomness inherent in the Bidirectional Rapidly-exploring Random Tree (RRT) algorithm, this paper proposes an improved Unmanned Aerial Vehicle (UAV) flight path planning strategy based on a goal-biased Bidirectional Artificial Potential Field method RRT (BI-APF-RRT) algorithm. Firstly, a goal-biasing strategy is employed for the generation of random samples and to guide the expansion direction of the bidirectional trees. The search is completed by establishing two alternately growing random search trees via the bidirectional RRT mechanism, thereby enhancing the algorithm’s convergence rate. Secondly, an improved Artificial Potential Field (APF) method is integrated into the bidirectional tree growth process, which further significantly reduces the number of search iterations. Concurrently, during the path smoothing phase, trajectory optimization is achieved through the application of a cubic B-spline interpolation algorithm. Finally, simulation experiments demonstrate that, compared to several existing algorithms, the proposed goal-biased BI-APF-RRT algorithm effectively reduces the number of iterations, enhances convergence speed, improves the directionality of new node generation, and effectively lowers the path cost. Consequently, it successfully mitigates the significant search space randomness problem associated with the conventional bidirectional RRT algorithm.