基于视觉阻抗控制的并联机械臂锥孔装配

Research on Parallel Robot Hole Assembly Based on Visual-Impedance Control

针对当前协作机器人在锥孔装配中依赖准确工件位置信息的局限性问题,提出了一种基于视觉阻抗控制的并联机械臂控制方法。首先,根据机器人锥孔装配的原理建立其动力模型,采用针孔模型建立图像特征点的运动模型;然后,基于所构建的动力模型,设计了基于视觉伺服控制和阻抗控制的机械臂混合控制策略,通过末端搭载的视觉系统获取的工件之间的特征点误差以实现机械臂末端的位置和力控制,从而完成工件之间的装配动作;最后,建立了机械臂系统实验平台,进行了工件装配的位置跟踪和末端接触力实验,并将其与传统的力位控制方法进行了对比实验,结果验证了该控制方法的有效性和稳定性。

In response to the limitations of current collaborative robots relying on precise workpiece position information in cone-hole assembly,we propose a parallel robot control method based on visual impedance control.Firstly,we establish a dynamic model for robot cone-hole assembly based on the principles of robot assembly and create a motion model for image feature points using the pinhole model.Then,based on the constructed dynamic model,we design a hybrid robot control strategy combining visual servo control and impedance control.This strategy utilizes the feature point error between workpieces obtained by the visual system mounted at the robot′s end effector to achieve both position and force control at the robot′s end effector,thereby completing the assembly between workpieces.Finally,a mechanical arm system experimental platform was established,and position tracking and end-effector contact force experiments for workpiece assembly were conducted.A comparative experiment was also conducted with traditional force-position control methods,validating the effectiveness and stability of this control approach.