李洋, 徐达. 基于引力自适应步长RRT的双臂机器人协同路径规划[J]. 机器人, 2020, 42(5): 606-616. DOI: 10.13973/j.cnki.robot.190592
引用本文: 李洋, 徐达. 基于引力自适应步长RRT的双臂机器人协同路径规划[J]. 机器人, 2020, 42(5): 606-616. DOI: 10.13973/j.cnki.robot.190592
LI Yang, XU Da. Cooperative Path Planning of Dual-arm Robot Based on Attractive Force Self-adaptive Step Size RRT[J]. ROBOT, 2020, 42(5): 606-616. DOI: 10.13973/j.cnki.robot.190592
Citation: LI Yang, XU Da. Cooperative Path Planning of Dual-arm Robot Based on Attractive Force Self-adaptive Step Size RRT[J]. ROBOT, 2020, 42(5): 606-616. DOI: 10.13973/j.cnki.robot.190592

基于引力自适应步长RRT的双臂机器人协同路径规划

Cooperative Path Planning of Dual-arm Robot Based on Attractive Force Self-adaptive Step Size RRT

  • 摘要: 快速扩展随机树(RRT)方法的步长确定过分依赖于程序调试,而且固定的步长会导致碰撞检测失效问题.针对此问题,本文提出一种适用于双臂机器人协同路径规划的引力自适应步长RRT.首先,通过建立构型空间与工作空间的步长范数不等式,对双臂机器人在工作空间中所产生的步长进行约束,进而确保实现有效的碰撞检测;然后,提出随机树被动生长方法,在保证双臂机器人协同运动的基础上,降低规划空间的维度.最后,在随机树的节点处引入引力函数,加快算法的融合速度.仿真结果表明,引力自适应步长RRT方法可对工作空间中的步长进行有效约束,确保算法碰撞检测的有效性.在无碰撞的前提下,引力自适应步长RRT方法相比于其他算法减少了迭代次数,降低了运行时间并缩短了路径长度.将所提算法应用于双臂机器人的样机实验,结果表明双臂机器人可在保持位置协同的前提下,完成避障运动,验证了算法的可行性.

     

    Abstract: In the RRT (rapidly-exploring random tree) method, the determination of the step size depends too much on program debugging, and collision detection failure may occur due to the fixed step size. For this problem, an attractive adaptive step size RRT for cooperative path planning of the dual-arm robot is proposed. Firstly, the step size norm inequality between the configuration space and the workspace is established to constrain the step size generated by the dual-arm robot in the workspace, and thus the effective collision detection is guaranteed. Then, a passive growth method of random tree is proposed to reduce the dimension of planning space while ensuring cooperative motion of the dual-arm robot. Finally, the attractive force function is introduced at the nodes of the random tree to speed up the fusion of the algorithm. The simulation results show that the attractive force self-adaptive step size RRT method can constrain the step size in the workspace effectively to ensure the effectiveness of collision detection. On the premise of no collision, the attractive adaptive step size RRT method reduces the number of iterations, the running time and the path length compared with other algorithms. The proposed algorithm is applied to the prototype experiment of the dual-arm robot. The experiment results show that the dual-arm robot can complete the obstacle avoidance motion on the premise of maintaining the position coordination, which verifies the effectiveness of the algorithm.

     

/

返回文章
返回