电解槽样品转运机器人的视觉伺服运动控制方法  

Visual Servo Motion Control Method for Electrolyzer Sample Transfer Robot

作　　者：王镕涛 樊绍胜[1] WANG Rongtao;FAN Shaosheng(School of Electrical and Information Engineering,Changsha University of Science and Technology,Changsha,Hunan 410114,China)

机构地区：[1]长沙理工大学电气与信息工程学院,湖南长沙410114

出　　处：《自动化应用》2023年第11期13-16,共4页Automation Application

摘　　要：针对电解槽样品转运机器人无标定视觉伺服控制对图像雅可比矩阵存在估计不准确、估计时间长的问题,本文提出了一种基于BP神经网络的KF-BP算法。KF-BP算法使用BP神经网络训练样本,输出误差补偿量,与传统KF算法产生的次优估计值相加得到雅可比矩阵的最优估计值,有效提高了雅可比矩阵估计的准确性和快速性。本文建立了基于KF-BP算法的无标定视觉伺服模型,并进行了仿真实验,结果表明,与传统KF算法相比,基于KF-BP算法的视觉伺服收敛速度提高了34.7%,且图像累计误差更小,可有效提高电解槽铝样品运机器人的作业精度和速度。A KF-BP algorithm based on BP neural network is proposed to address the problem of inaccurate estimation and long estimation time of image Jacobian matrix in uncalibrated visual servo control of electrolytic cell sample transfer robots.The KF-BP algorithm uses a BP neural network to train samples,outputs an error compensation,and adds it to the suboptimal estimates generated by the traditional KF algorithm to obtain the optimal estimate of the Jacobian matrix,effectively improving the accuracy and speed of Jacobian matrix estimation.This paper establishes an uncalibrated visual servo model based on the KF-BP algorithm and conducts simulation experiments.The results show that compared with the traditional KF algorithm,the convergence speed of the visual servo based on the KF-BP algorithm is improved by 34.7%,and the cumulative error of the image is smaller.This can effectively improve the operation accuracy and speed of the aluminum sample transportation robot in the electrolytic cell.

关 键 词：电解槽样品转运机器人 无标定视觉伺服 KF-BP算法 BP神经网络 

分 类 号：TP242[自动化与计算机技术—检测技术与自动化装置]