基于变刚度自适应导纳机制的机械臂恒力控制

工业机械臂在诸如打磨抛光等接触式作业任务中对环境刚度信息存在一定的依赖性, 未知环境刚度信息将严重影响机器人的力位控制精度, 使得作业效果难以得到保证. 为解决环境信息不足或未知情况下的力/位置精确控制问题, 本文首先提出了一种新的自适应环境刚度在线估计方法, 针对时变的环境刚度进行实时估计, 由此预测生成后继的机械臂参考轨迹点, 随后提出了一种根据力跟踪误差实时调整末端工具手刚度系数的变刚度导纳恒力控制方法, 并结合李雅普诺夫稳定性理论给出了整体控制律的收敛性证明. 针对刚柔两种末端工具手和多种不同的曲面工件开展了实验研究, 并与传统PID控制方法和传统导纳控制方法进行了对比, 其结果表明本文所提出的复合控制方法可在不同工况条件下实现机器人运动过程中接触力的快速柔顺调节, 并获得4.55%以内的最优力控误差效果, 证明了本文所提出方法的有效性与可行性.

Constant force control of manipulator based on variable stiffness adaptive admittance mechanism

The industrial manipulator has a certain dependence on the environmental stiffness information in contact tasks such as grinding and polishing. The unknown environmental stiffness information will seriously affect the force and position control accuracy of the robot, making it difficult to guarantee the operation effect. In order to solve the problem of precise force/position control in the case of insufficient or unknown environmental information, a new online estimation method of adaptive environment stiffness is firstly proposed in this paper, which can estimate the time-varying environmental stiffness in real time, and then predict and generate subsequent reference trajectory points of the manipulator. Then, a variable stiffness admittance constant force control method is proposed to adjust the stiffness coefficient of the end tool hand in real time according to the force tracking error, and the convergence proof of the overall control law is given combined with the Lyapunov stability theory. The experimental research is carried out on two kinds of rigid and flexible end tool hands and a variety of different surface workpieces, and compared with the traditional PID control method and the traditional admittance control method. The results show that the composite control method proposed in this paper can be used in different work The rapid and compliant adjustment of the contact force during the robot motion is realized under the condition of the robot, and the optimal force control error effect within 4.55% is obtained, which proves the effectiveness and feasibility of the method proposed in this paper.