摘要
为研究六轴机器人控制坐标从球面叠加坐标系向三维直角做标记进行机器学习坐标转换的核心控制算法,将机器学习算法引入六轴机器人控制为核心思路,通过有外部变量干预的多列神经网络控制系统革新六轴机器人自带控制系统,实现系统内球面坐标控制模式向三维直角坐标控制模式的革新。当达到100000次训练后,基于神经网络的革新控制算法可以实现自主控制机器人操作部件控制中心点几乎沿空间最短直线运动,其控制效率远大于机器人自带控制系统。证明该革新控制系统对六轴机器人的自主控制有积极意义。
To study the core control algorithm of six axis robot control coordinate transformation from spherical superposition coordinate system to three-dimensional right angle marking,the machine learning algorithm was introduced into the six axis robot control as the core idea,and the six axis robot’s own control system was innovated through the multi column neural network control system with the intervention of external variables in order to achieve the innovation from the spherical coordinate control mode in the system to the three-dimensional rectangular coordinate control mode.After 100000 times of training,the autonomous control of the robot operating parts was realized by the innovative control algorithm based on neural network,with the control center point moving almost along the shortest straight line in space.The results show that the new control system is much more efficient than that of the robots of its own and has positive significance for the autonomous control of six axis robot.
作者
何雅芹
HE Yaqin(Langfang Central Blood Station,Langfang 065000,China)
出处
《机械制造与自动化》
2021年第3期213-215,239,共4页
Machine Building & Automation
关键词
六轴机器人
球面叠加坐标系
三维直角坐标控制模式
坐标转换算法
six axis robot
spherical superposition coordinate system
three-dimensional rectangular coordinate system
coordinate transformation algorithm
