Research on trajectory planning of coal gangue sorting robot based on APF-TB-RRT* algorithm
MIN Zhenhui
WANG Zheng
WANG Yanwei
ZHANG Chenyang
Abstract：[Objective]Aiming at the defects of long computation time,redundant paths and poor smoothness of the traditional rapidly-exploring random tree star(RRT*)algorithm in obstacle avoidance trajectory planning of coal gangue sorting manipulators,as well as the problems of large modeling redundancy of irregular gangue by the existing spherical bounding box method and easy falling into local optimum by the fixed target bias strategy,an improved APF-TB-RRT* algorithm integrating the axis-aligned bounding box method,artificial potential field(APF)method and target bias(TB)strategy was proposed.[Methods]Firstly,the axis-aligned bounding box was adopted to simplify the modeling of irregular gangue obstacles to reduce the computational cost of collision detection.Secondly,the core parameters of the algorithm were optimized through a 3-factor 6-level orthogonal test,and a three-probability hybrid sampling mechanism was designed.The artificial potential field method was selected probabilistically to generate a repulsive force field within the bounding box to guide path growth,and the target region sampling was selected probabilistically to balance the algorithm guidance and randomness.Then,the cubic B-spline curve was used to interpolate and optimize the initial path to improve trajectory smoothness.Finally,the algorithm performance was verified through Matlab software simulation and prototype tests.[Results]The results show that the trajectory planning success rate of the improved algorithm reaches 100%.In obstacle-free,simple and complex obstacle environments,the running time is shortened by 97.07%,63.19%and 75.95%respectively compared with the RRT* algorithm with fixed target bias strategy;the maximum reduction of path length variance is 86.12%,and the algorithm stability is significantly improved.Prototype tests show that the obstacle-avoidance path planned by the improved algorithm shortens the overall path length by 9.25%compared with the traditional"vertical up-and-down"path,and the end-effector of the manipulator can accurately track the planned trajectory.
Keywords：Coal gangue sortingRobotic armTrajectory planningRapidly-exploring random treeArtificial potential fieldBounding box method
Publication Date：2026-06-30
Online Publishing Date：2026-09-12(First online date of this platform, not the publication date of the document)
Pages：15( 146-160 )