基于期望动力学的柔性关节控制器设计

协作机器人具有灵活，安全特点，已广泛应用于自动化领域以及中小企业中。为了保证与人交互的安全性，协作机器人通常采用中空电机与中空减速器配合的设计方案，以降低关节转动惯量，从而获得良好的外力感知与控制能力，这种设计导致协作机器人的关节具有了柔性。针对具有柔性关节的轻量级协作机器人，设计了一种基于期望动力学的柔性关节控制器，提高柔性关节机器人的轨迹跟踪精度和抖动抑制能力。在具有谐波减速器和力矩传感器的柔性关节上，基于连杆侧位置反馈与关节力矩反馈实现了从经典的电机侧控制到连杆侧控制的转变，并借助储存函数建立李雅普诺夫函数证明了该控制器的无源性与渐近稳定性。最后，通过Simulink仿真与单关节实验平台的关节轨迹追踪实验验证了柔性关节控制器的性能，结果显示其与全状态反馈控制相比具有关节力矩波动小、抖动抑制快以及轨迹跟踪误差小等优点。

Design of Flexible Joint Controller Based on Desired Dynamics

Cooperative robots are flexible and safe, and have been widely used in automation and small and medium-sized enterprises. In order to ensure the safety of interaction with people, cooperative robots usually use the design scheme of hollow motor and hollow reducer to reduce the joint moment of inertia, so as to obtain good external force perception and control capability. This design leads to the flexibility of the joint of cooperative robots. A flexible joint controller based on expected dynamics is designed for a cooperative robot with flexible joints, which improves the trajectory tracking accuracy and vibration suppression ability of the flexible joint robot. On the flexible joint with harmonic reducer and torque sensor, the transformation from classical motor side control to link side control is realized based on link side position feedback and joint torque feedback. Further, the storage function is used to establish the Lyapunov function, and the passivity and asymptotic stability of the controller are proved. Finally, the performance of the flexible joint controller is verified through Simulink simulation and joint trajectory tracking experiment on a single joint experimental platform. The results show that compared with the full state feedback control, it has the advantages of small joint torque fluctuation, faster vibration suppression and smaller trajectory tracking error.