论文部分内容阅读
机械臂运动规划是机器人研究领域的重点,对机械臂能否顺利执行任务非常重要。目前,机械臂运动规划多使用RRT法,然而该方法是在关节空间进行规划,无法适用于机械臂末端执行器存在约束的任务。为了克服这个不足,该文提出了一种任务自由子空间RRT(rapidly-exploring random tree)法,在末端执行器任务空间的自由子空间中构建RRT,并对其每步扩张进行逆运动学轨迹优化,求解出相应的关节轨迹。此外,由于末端执行器速度对逆运动规划有重要影响,该文在逆运动轨迹优化阶段采用了最似梯度法,不仅考虑了末端执行器的运动速度,而且通过极小关节自由速度和优化目标负梯度的距离,重新确定了关节自由速度,增强了算法的优化能力。实验结果表明:该算法能有效解决机械臂末端存在约束的问题。
Robotic arm planning is the focus of robotic research and is very important for the successful execution of robotic arm. At present, the RRT method is mostly used in the motion planning of the manipulator. However, this method is planned in the joint space and can not be applied to the constraint that the end effector of the manipulator has a constraint. In order to overcome this shortcoming, this paper presents a fast-exploring random tree (RRT) method, which constructs the RRT in the free subspace of the end effector task space and performs inverse kinematics trajectory Optimize, solve the corresponding joint trajectory. In addition, since the velocity of the end effector has an important influence on the inverse motion planning, the most gradient-like method is adopted in the reverse trajectory optimization stage, which not only considers the velocity of the end effector, but also minimizes the velocity of the joint and optimizes the target Negative gradient distance, to re-determine the freedom of joint speed and enhance the algorithm’s ability to optimize. Experimental results show that the proposed algorithm can effectively solve the problem of constraint at the end of manipulator.