改进的双向RRT路径规划算法研究
Research on Improved Bidirectional RRT Path Planning Algorithm

针对传统BI-RRT算法在路径规划中双向扩展树扩展过程中无目标偏向性，收敛速度慢，易出现冗余点问题，提出了一种修剪STB-RRT算法。该算法在随机节点采样中加入偏向函数，通过对节点的处理，使随机树的生长更具有方向性，然后再加入删除冗余点的修剪策略，进一步提高收敛速度，最后通过阈值判断两棵随机树是否连接。通过在两种环境下进行实验仿真，证实该改进算法可以明显提升收敛速度，减少规划路径长度，具有较高的实用性和有效性。
A pruning STB-RRT algorithm is proposed for the traditional BI-RRT algorithm with no target bias in the process of bi-directional expansion tree expansion in path planning, slow convergence speed and easy-to-appear redundant points. The algorithm adds a bias function to the random node sampling to make the random tree grow more directional by processing the nodes, then adds a pruning strategy to remove redundant points to further improve the convergence speed, and finally judges whether the two random trees are connected by a threshold value. Through experimental simulations in both environments, it is confirmed that the improved algorithm can significantly improve the convergence speed and reduce the planning path length, which has high practicality and effectiveness.