Path Planning for a Prostate Intervention Robot Based on an Improved Bi-RRT Algorithm

In the medical field, the application of robots must primarily ensure safety. An improved path-planning method based on bidirectional rapidly exploring random tree (Bi-RRT) is proposed to enable prostate intervention robots to avoid obstacles in unstructured natural environments. To address the limitations of the existing Bi-RRT algorithm, in our method, we introduce a greedy expansion strategy with dynamic nodes as the target and add a start point repulsion function to the random sampling process as well as sampling space limitations, resulting in a significant improvement in the sampling speed and efficiency. The RRT* parent node reselection and rewiring strategy is introduced to replace the original method to reduce the path cost. To remove redundant nodes and generate a smooth, curved, and continuous robot-executable path, a path optimization strategy based on a fusion of pruning functions, isometric interpolation, and cubic B-spline interpolation is proposed. Finally, a comparative simulation study of the different algorithms using MATLAB and realistic robot obstacle avoidance experiments in an unstructured natural environment is carried out to verify the effectiveness of the proposed method. The method can boost the speed of path planning, decrease the path cost, and create smooth paths that avoid obstacles.