Improved Rapid-Expanding-Random-Tree-Based Trajectory Planning on Drill ARM of Anchor Drilling Robots

Permanent highway support in deep coal mines now depends on the anchor drilling robot's drill arm. The drilling arm's trajectory planning using the conventional RRT (rapid-expanding random tree) algorithm is inefficient and has crooked, rough paths. To improve the accuracy of path planning, we propose an improved RRT algorithm. Firstly, the kinematic model of the drill arm of the drill and anchor robot was established, and the improved DH solution parameters and the positive solution of the drill arm kinematics were solved. The end effector's attainable working space was calculated using the Monte Carlo approach. Additionally, to address the problem of the slow running speed of the RRT algorithm, an artificial potential field factor was introduced to construct virtual force fields at obstacle and target points and calculate the potential field map for the entire reachable workspace to improve the speed of the sampling points close to the target point. At the same time, the greedy approach and the three-time B-sample curve-fitting method were used simultaneously to remove unnecessary points and carry out smooth path processing in order to improve the quality of the drill arm trajectory. This was carried out in order to solve the issue of rough pathways generated by the RRT algorithm. Finally, 50 time-sampling comparison experiments were conducted on 2D and 3D maps. The experimental results showed that the improved RRT algorithm improved the average sampling speed by 20% and reduced the average path length by 14% compared with the RRT algorithm, which verified the feasibility and effectiveness of this improved RRT algorithm. The improved RRT algorithm generates more efficient and smoother paths, which can improve the intelligence of the support process by integrating and automating drilling and anchoring and providing reliable support for coal mine intelligence.