采用改进RRT方法进行双臂机器人路径规划

Path Planning Method of Dual-arm Robot Based on Improved RRT

单臂机器人只适用于单一、简单的应用场景,存在很大的局限性。双臂机器人操作灵活性高,交互能力强,能够适应复杂的应用场景。但是,双臂机器人的路径规划是一个技术难点,不仅需要考虑机器人与环境的避障问题,还需考虑双臂之间的避碰问题。针对上述问题,提出了面向双臂机器人的改进RRT(Rapidly-ExploringRandom-Tree)方法,在关节空间中对机械臂进行路径规划,减少了计算次数,提高了计算效率。针对探索树随机生长没有方向的缺点,加入引力势场,给随机生长树的探索加入了方向,加快其运算速度,进一步提高了效率。针对RRT规划出的路径曲线曲折的缺点,采用贝塞尔曲线进行平滑处理,使机械臂运动更加平稳。采用主从协调的方法,先对主臂进行避障路径规划,后依据主臂的路径对从臂进行避障和避碰路径规划。通过MATLAB仿真和样机实验验证了算法的可行性。

Single-arm robots are only suitable for single and simple application scenarios,which have great limitations.The dual-arm robots have high operational flexibility and strong interaction capabilities,and can adapt to complex application scenarios.However,path planning for a dual-arm robot is difficult.During the path planning,there are two problems that should be considered.The first one is the obstacle avoidance between the robot and the environment,the other is the collision avoidance between the two arms.Based on these problems,this paper proposes an improved rapidly-exploring random-Tree(RRT)method for dual-arm robots.This method is applied in the joint space,which reduces the number of calculation and improves efficiency.Moreover,the traditional RRT has some shortcomings.First of all,the random growth of the exploration tree has no direction.Therefore,the gravitational potential field is added to add a direction to the exploration tree.This method can speed up its calculation speed and improves the efficiency.Secondly,the path curve planned by RRT is tortuous.Hence,the Bezier curve is used to smooth it to make the movement of the manipulator more stable.The master-slave coordination method is adopted to plan the dual arm manipulator.Firstly,the obstacle avoidance path of the master arm is calculated.And then obstacle avoidance and collision avoidance paths of the slave arm are determined according to the path of the master arm.The feasibility of the algorithm is verified by MATLAB simulation and prototype experiment.