摘要
为提高工业机器人绝对定位精度,提出一种基于DBO-BP与离线前馈校正相结合的方法。该方法适用于工业机器人定位误差补偿研究。通过使用拉丁超立方抽样法获取工业机器人的位姿样本,并利用BP神经网络建立误差预测模型,应用DBO优化算法改善了局部最优现象,从而提高了模型的收敛性和鲁棒性。经过离线前馈补偿处理后,降低了工业机器人定位误差,大幅提高了机器人绝对定位精度。这种方法能够有效提高机器人的精度和稳定性,并为工业机器人的精准定位问题提供了可行的解决方案。
In order to improve the absolute positioning accuracy of industrial robots,a method based on DBO-BP and offline feedforward correction was proposed.This method is suitable for the research on positioning error compensation of industrial robots.By using Latin Hypercube Sampling method to obtain the pose samples of industrial robots,and using BP neural network to establish an error prediction model,the DBO optimization algorithm was applied to improve the local optimal phenomenon,thus improving the convergence and robustness of the model.After offline feedforward compensation processing,the positioning error of industrial robots was reduced,and the absolute positioning accuracy of robots was greatly improved.This method can effectively improve the accuracy and stability of robots,and provides a feasible solution for the precise positioning problem of industrial robots.
作者
刘麒
谭丁诚
刘振刚
王影
LIU Qi;TAN Dingcheng;LIU Zhengang;WANG Ying(School of Information and Control Engineering,Jilin Institute of Chemical Technology,Jilin City 132022,China;Staubli(Hangzhou)Precision Machinery&Electronics Co.,Ltd.Hangzhou 310002,China)
出处
《吉林化工学院学报》
CAS
2024年第1期59-66,共8页
Journal of Jilin Institute of Chemical Technology
关键词
工业机器人
BP神经网络
DBO算法
绝对定位精度
误差补偿
industrial robots
BP neural network
DBO algorithm
absolute positioning accuracy
error compensation
