摘要
工业机器人末端抓取操作器面对具有几何不变性特性的B样条曲线轨迹时,控制精度较差,为此提出一种工业机器人复杂B样条曲线轨迹控制精度补偿方法。根据工业机器人关节转角和偏距等建立运动学模型,分析导致复杂B样条曲线轨迹误差的主要因素,采用笛卡儿坐标系和线性弹簧模型描述几何参数和柔性误差。以工业机器人实际轨迹作为输入,高斯函数作为径向基函数,构建径向基神经网络模型,选择合理的基函数中心,通过学习得到工业机器人复杂B样条曲线轨迹控制精度补偿结果。实验结果表明:所提方法的机器人实际轨迹与期望轨迹重合度较高,最高轨迹控制误差仅为0.76 mm,执行时间最高仅为1.9 s,有效补偿了控制精度。
To improve the poor control accuracy of the end grabbing manipulator of an industrial robot when facing a B-spline curve trajectory with geometric invariance,a compensation method for the control accuracy of complex B-spline curve trajectory of an industrial robot is proposed.Based on the joint angle and offset of the industrial robot,the kinematics model is established,and the main factors leading to the trajectory error of the complex B-spline curve are analyzed.The geometric parameters and flexible errors are described by Cartesian coordinate system and linear spring model.With the actual trajectory of the industrial robot as the input and the Gaussian function as the radial basis function,the radial basis function neural network model is constructed,and the reasonable basis function center is selected.The control accuracy compensation result of the complex B-spline curve trajectory of the industrial robot is obtained through learning.The experiment results show that the proposed method has the highest coincidence between the actual trajectory and the expected trajectory,the highest trajectory control error is only 0.76mm,and the highest execution time is only 1.9s,which effectively compensates the control accuracy.
作者
颜双权
胥建成
YAN Shuangquan;XU Jiancheng(Department of Intelligent Engineering,Liaoning Institute of Science Engineering,Jinzhou 121010,China)
出处
《机械制造与自动化》
2023年第5期32-35,共4页
Machine Building & Automation
关键词
径向基神经网络
工业机器人
运动轨迹
精度补偿
高斯函数
B样条曲线
radial basis function neural network
industrial robot
motion trajectory
accuracy compensation
Gaussian function
B-spline curve