摘要
本文提出一种无外力传感器、稳定高效的基于关节电机速度控制的机器人直接示教策略。对示教过程中的机器人重力、摩擦力分别进行建模分析。通过机器人的力/力矩平衡方程,递推作用于关节上的重力,并在MATLAB和Adams中仿真验证。建立包括库伦摩擦力和粘性摩擦力的摩擦力模型,并进行模型参数识别。最后,在UR10机器人测试平台上进行了控制策略实现和直接示教实验,实验示教效果显著。该控制策略运算量小、系统响应快、示教灵活、成本低,为后续机器人直接示教研究提供新的方法和途径。
This paper proposes a stable and efficient robot direct teaching strategy based on joint motor speed control without external force sensor. The gravity and friction of the robot during the teaching process were modeled and analyzed respectively. Through the force/torque balance equation of the robot,the gravity acting on the joint is recursively verified by simulation in MATLAB and Adams. A friction model including Coulomb friction and viscous friction is established,and the model parameters are identified. Finally,the control strategy implementation and direct teaching experiments were performed on the UR10 robot test platform,and the experimental teaching effect was remarkable. The control strategy has a small amount of calculations,a fast system response,flexible teaching,and low cost,and provides a new method and approach for subsequent robot direct teaching research.
作者
文鑫
庄正浩
WEN Xin;ZHUANG Zhenghao
出处
《计量与测试技术》
2020年第4期28-32,36,共6页
Metrology & Measurement Technique
