机器人几何参数误差识别与补偿方法及实验

Robot geometric parameter error identification and compensation method and experiment

  • 摘要: 工业机器人因其组织灵活性、强可拓展性,在叶片类复杂零件的高精度加工中得到越来越多的关注,这对机器人末端执行机构的定位精度提出了更高的要求。文章基于自主设计研发的六自由度关节机器人进行几何参数误差识别和补偿方法研究,旨在提高机器人末端执行机构的定位精度。首先,基于D-H法建立机器人各关节转角与末端位置坐标的关系模型,建立分别考虑连杆长度误差、连杆偏置误差和关节转角误差的位置误差模型;然后,基于自主设计的专用测头进行运动过程坐标数据测量实验,基于实验数据采用Levenberg-Marquardt进行几何参数误差项识别;最后,将识别的结果在机器人模型中进行补偿,并对补偿后的效果进行实验验证,结果表明,补偿后的定位精度比补偿前提高了60%以上。

     

    Abstract: Industrial robots are getting more and more attention in the high-precision machining of complex parts of blade type due to their organizational flexibility and strong expandability, which puts forward higher requirements on the positioning accuracy of the robot end-actuator. In this paper, based on the six-degree-of-freedom articulated robot designed and developed independently, we conduct a research on the identification of geometric parameter errors and compensation methods, aiming to improve the positioning accuracy of the robot end-actuator. Firstly, the relationship model between the robot’s joint angles and end position coordinates is established based on the D-H method, and the position error model is established by considering the connecting rod length error, connecting rod bias error and joint angle error respectively. Secondly, the coordinate data measurement experiment is carried out based on the independently-designed special probe for the kinematic process, and the geometric parameter error term identification is carried out by using the Levenberg-Marquardt based on the experimental data. Finally, the geometric parameter error term identification and compensation method is carried out based on the independently-designed special probe error term identification. The identified results are compensated in the robot model and the effect of the compensation is experimentally verified, and the results show that the positioning accuracy after compensation is improved by more than 60% compared with the pre-compensation.

     

/

返回文章
返回