【摘要】 为解决六轴工业机器人关节柔性引起的末端振动问题，提出一种基于力矩输入整形补偿的抑制方法。建立柔性关节伺服控制模型，分析关节振动成因。基于此，在柔性关节前引入最优输入整形器，通过对电机输出力矩进行输入整形后提取波动力矩，并将其以力矩补偿的形式反馈至电流环前，以迅速平衡关节弹性力矩变化并滤除电机力矩在自然振动频率点附近的频率。在Simulink中搭建机器人关节2仿真模型，对比补偿前后位置、速度、加速度响应。最后，为验证所提方法在六轴机器人系统中的可行性，搭建工业机器人实验平台，对比力矩输入整形补偿前后位置跟踪误差。结果表明：机器人运动过程中，电机输出力矩未能及时平衡关节弹性力矩变化是导致关节振动的主要原因；采用力矩输入整形补偿方法后，振动幅度和持续时间明显下降，1.1 s后关节位置、速度和加速度振动信号几乎完全被滤除；使用该方法后，各关节位置跟踪误差均方根显著降低，关节1～6分别减少58.24%、58.79%、59.57%、54.37%、63.39%和59.8%,验证了力矩输入整形补偿方法在机器人振动抑制中的可行性和有效性。更多还原

【Abstract】 To address the end-effector vibration issue caused by joint flexibility in six-axis industrial robots, a suppression method based on torque input shaping compensation was proposed.A servo control model for flexible joints was established, and the causes of joint vibration were analyzed.Based on this, an optimal input shaper was introduced before the flexible joint.After input shaping the motor output torque, the fluctuating torque was extracted and fed back to the front of the current loop in the form of torque compensation, so as to quickly balance the change of joint elastic torque and filter out the frequencies of motor torque near the natural vibration frequency point.A simulation model for joint 2 of the robot was built in Simulink, and the position, velocity, and acceleration responses before and after compensation were compared.Finally, to verify the feasibility of the proposed method in the six-axis robot system, an industrial robot experimental platform was built, and the position tracking errors before and after torque input shaping compensation were compared.The results show that during the robot′s movement, the failure of the motor output torque to balance the change of joint elastic torque in a timely manner is the main cause of joint vibration.After adopting the torque input shaping compensation method, the vibration amplitude and duration are significantly reduced, and the vibration signals of joint position, velocity, and acceleration are almost completely filtered out after 1.1 s.After using this method, the root-mean-square of position tracking errors of each joint is significantly reduced, with joint 1 to joint 6 decreasing by 58.24%,58.79%,59.57%,54.37%,63.39%,and 59.8% respectively, which verifies the feasibility and effectiveness of the torque input shaping compensation method in robot vibration suppression.更多还原