This paper discusses the motion control problem of kinematically redundant hybrid parallel robots that were recently proposed. The kinematic and dynamic models are firstly reviewed. It is pointed out that the robot can be decomposed into two parts and that each part can be analysed independently. A new hybrid approach is proposed based on this property of the robot. This approach includes an adapted computed-torque control scheme for the legs that operates in the joint space as well as a compensation of the errors of the platform applied in the Cartesian space. The convergence of this proposed approach is also verified using the Lyapunov stability theory. Two example architectures of kinematically redundant robots are built and experiments are conducted. Finally, the results are compared and analysed in order to validate the improvements provided by the proposed control method. It is shown that the proposed control scheme significantly reduces the position error. Potential extensions of this work are also discussed.

本文讨论了最近提出的运动学冗余混合并联机器人的运动控制问题。首先回顾了运动学和动力学模型。要指出的是，机器人可以分解为两个部分，并且每个部分都可以独立分析。基于机器人的这一特性，提出了一种新的混合方法。该方法包括针对在关节空间中操作的腿部的自适应计算扭矩控制方案，以及对在笛卡尔空间中应用的平台的误差进行补偿。还使用李雅普诺夫稳定性理论验证了该方法的收敛性。建立了运动学冗余机器人的两个示例架构，并进行了实验。最后，为了验证所提出的控制方法所提供的改进，对结果进行了比较和分析。结果表明，所提出的控制方案大大降低了位置误差。还讨论了这项工作的潜在扩展。