基于位置控制的多自由度机械臂非线性比例-微分阻抗控制

机器人柔顺控制可以响应环境变化，但接触信息的延迟以及未知机器人系统的跟踪误差等问题均导致接触瞬间力矩超调严重。针对上述问题本文提出一种基于自适应位置控制的改进阻抗控制策略，实现快速、精确的位置跟踪，同时，提高力控制的响应速度和精度。本策略采用双环控制，外环在传统阻抗模型基础上引入非线性接触力微分项在保持系统稳定性的同时提高机器人对接触力变化的响应，有效降低接触力超调；内环为自适应滑模控制，并使用RBF神经网络逼近机器人动力学模型并补偿系统中不确定性扰动，提高了系统的鲁棒性，提高收敛速度并降低跟随误差。通过仿真与实验，验证了所提出的改进阻抗控制方法相比于传统的阻抗控制方法有更好的力控响应速度和位置跟踪精度，可有效解决机器人与环境接触瞬间的接触力超调问题。

Position-Based Impedance Control of a Multi-DOF Arm by Using Nonlinear Proportional-Differential Operation

Robot compliant control is able to respond to environmental changes, but issues such as delayed contact information and tracking errors of unknown robot systems lead to significant overshoot of contact instantaneous torque. To address these concerns, this study proposes an improved impedance control strategy based on adaptive position control to achieve rapid and accurate position tracking, while enhancing the response speed and accuracy of force control. This strategy employs a dual-loop control, in which the outer loop incorporates a non-linear contact force differential term on the foundation of the conventional impedance model to maintain system stability while improving the robot''s response to contact force changes, thereby effectively reducing contact force overshoot. The inner loop is an adaptive sliding mode control, which leverages RBF neural network to approximate the robot''s dynamic model and compensate for the uncertainty disturbance in the system, thus increasing the system''s robustness, enhancing convergence speed, and decreasing tracking errors. Through simulation and experimentation, the proposed improved impedance control method is confirmed to possess better force control response speed and position tracking accuracy than the traditional impedance control method, and can effectively address the issue of contact force overshoot during robot-environment contact.