摘要
本文讨论了受控制驱动力、外界扰动及参数不确定的一种舵机组合控制策略的书写机器人设计。首先,构建多舵机组合机械臂模型及其旋转拟合空间,解算并采集其正、逆运动学和动力学的仿真样本数据训练RBF神经网络自适应策略控制器,分析多舵机组合控制后的关节转角轨迹曲线及系统响应特性;进而,针对书写机器人组合舵机的运行轨迹目标跟踪末端书写过程,完成4自由度和6自由度舵机关节三维空间坐标的建模和仿真,采用RBF神经网络自适应策略和PNN神经网络策略完成书写机器人末端位置控制,比较两种控制策略的书写机器人末端控制过程的位置误差精度参数。研究表明,根据舵机组合控制策略进行书写机器人的建模与分析的特征参数,进行设计相应的书写机器人控制系统,笔尖末端轨迹运行控制能够5s内执行上位机指令,文中设计的硬笔书写机器人和软笔书写机器人都能写出具有良好辨识度的图形格式。
A writing robot design on multi -servo composite control strategy is discussed with the driving force, external disturbances and uncertain parameters. Firstly, the joint rotation trajectory curve and system response characteristics of the writing robot are produced from the RBF neural network, which the controller is trained by simulation sample data from the rotation combination fitting space, calculation and collection of the positive and inverse kinematics and dynamics. Then, comparing the position error precision parameters from terminal control process of two control strategies for writing robot, the RBF neural network and the PNN neural network are trained by simulation sample data from the Modeling of 3D coordinates of servo joints with four degrees or six degrees of freedom to show the target terminal tracking control process of a writing robot. The research indicated that the pen writing robot and brush writing robot designed in the paper could output good format graphics to be identified, the multi - servo composite control system of the writing robot could run and perform computer instruction in 5 s.
作者
许晓飞
Xu Xiaofei(School of Automation,Beijing Information Science and Technology University,Beijing 100192)
出处
《高技术通讯》
EI
CAS
北大核心
2018年第7期643-650,共8页
Chinese High Technology Letters
基金
北京信息科技大学教学改革(2016JGYB12)
北京市大学生科技创新(校教发[2018])资助项目
关键词
书写机器人
动态模型
舵机组合
RBF神经网络
writing robot
kinematic and dynamic model
a combined steering gear
RBF neural network