三平动力控末端执行器鲁棒自适应力跟踪导纳控制方法

针对现有重载工业机器人缺乏力控功能而难以满足去毛刺、倒角和磨抛等连续接触式作业要求的问题，提出一种含氮气弹簧的气电直驱式3-P(UU)2型三平动力控末端执行器，采用鲁棒自适应力跟踪导纳控制算法实现操作空间中接触面法向输出力的快速跟踪，适用于工件内外侧面、孔洞和狭小结构的机器人磨抛等过程。建立机构正逆运动学与气电直驱致动器动力学模型，并设计了鲁棒自适应力跟踪导纳控制器。实验结果表明，基于鲁棒自适应力跟踪导纳控制的三平动力控末端执行器力阶跃响应的稳态误差为-4.5×10^-4 N，上升时间19.27 ms，可实现力的快速精确跟踪；负载冲击下力的调整时间116.0 ms，最大超调量57.5%，具有良好的缓冲吸振特性与鲁棒性；在平面往复运动与圆柱面连续运动工况下，力均方根误差0.143 N，冲击峰值均值0.694N，对不同材质工具磨头的接触刚度变化与环境位移误差的适应性好，可提升工业机器人的连续接触式作业质量并拓宽其应用范围。

Robust Adaptive Force Tracking Admittance Control for 3-DOF Translational Force-controlled End-effector

The existing heavy-duty industrial robots are short of force control functionality, which makes them difficult to perform contact-operations such as deburring, chamfering, and polishing. A 3-DOF translational force-controlled end-effector featured with 3-P(UU)₂ parallel mechanism and gas springs inside is introduced, and the robust adaptive force tracking admittance control method is proposed to regulate the contact force in the operational space. It can be applied to the robotic grinding and polishing process of the lateral surfaces, holes and narrow structures of workpieces. Based on the modeling of the forward and inverse kinematics of the parallel mechanism, and the dynamics of the pneumoelectric actuator, the robust adaptive force tracking admittance controller is designed. Experimental results show that the 3-DOF translational force-controlled end-effector based on robust adaptive admittance control tracks force fast and precisely, with the steady-state force error ‒4.5×10^‒4 N and the rise time 19.27 ms. The settling time 116.0 ms and overshoot 57.5% demonstrates good impact resistance and vibration absorption characteristics. Furthermore, the standard deviation and average peak of the contact force during the movement in a plane and cylindrical surface are 0.143 N and 0.694 N, respectively, indicating the adaptability to the environmental displacement error and stiffness variation under grinding tools made of different materials. The add-on force-controlled end-effector presented can improve the quality of the advanced robotic contact operations and expand the applications of the industrial robots.