基于图像矩和矢量积法的六自由度机械臂视觉伺服控制

摘要

A visual servoing method based on image for a 6 DOF robot manipulator was proposed.On the basis of image moments and vector product,Jacobian matrix was introduced to resolve problems in robotic inverse kinematics by building robotic forward kinematics model and inverse kinematics model.Furthermore the composite Jacobian matrix was established,which stands for the mapping relationship between the image characteristic variation and joint angular velocity in Descartes space.The joint angular velocity of the manipulator in visual servoing was obtained from the variation of moments characteristic,ensuring that when the image approximates to the expected one,the manipulator can reach the desired place and the joint angular velocity converges to zero at the same time.The simulation results show that the method proposed to calculate the composite Jacobian matrix can control the manipulator to reach the expected position.%针对眼在手上的六自由度机械臂系统,提出一种基于图像的视觉伺服控制.通过图像矩和矢量积法,建立机器人正向、逆向运动学模型,引入雅可比矩阵解决机器人逆运动学解析问题.建立了图像矩特征变化量和笛卡尔空间的关节角速度之间的映射关系即复合雅可比矩阵,由矩特征变化量得到伺服过程中六自由度机器人各关节角速度.保证在图像逼近期望图像时,机械臂末端到达期望位置,并且此时关节角速度将收敛到零.仿真结果表明,计算复合雅可比矩阵的方法可控削机械臂渐近稳定到期望位置.