本发明公开了一种基于机器人关节动力学模型的机器人负载振动预测方法，属于机械振动及信号处理领域。本发明应用外部激励得到系统关于激励力和振动响应之间的传递函数，并且以此辨识得到关节中的未知参数，根据关节的各部分参数构建电机电磁转矩到手臂的传递函数，并且以此构建机器人手臂振动预测模型，根据该预测模型能高效预测振动情况，避免了在机器人上直接安装振动传感器来检测振动，经过实例求证，本发明可以有效地对工业机器人关节振动预测，具有广泛的实用性，而且预测结果可靠。

The invention discloses a robot load vibration prediction method based on a robot joint dynamics model, which belongs to the field of mechanical vibration and signal processing. According to the invention, the external excitation is applied to obtain the transfer function between the excitation force and the vibration response of the system, and unknown parameters in the joint are identified by this, and the transfer function from the electromagnetic torque of the motor to the arm is constructed according to the parameters of each part of the joint, and a robot arm vibration prediction model is constructed. According to this prediction model, the vibration situation can be predicted efficiently, and it is avoided that a vibration sensor is directly installed on the robot to detect the vibration. Through the verification of examples, the invention can effectively predict the joint vibration of industrial robots, and has wide practicability and reliable prediction results.