基于RBF神经网络的虚拟轴机床末端刀具运动位姿实时检测研究  

Real-time pose measurement for tool of a virtual axis machine tool based on a RBFNN

作　　者：高国琴[1] 薛娌[1] 张义贞[1] 

机构地区：[1]江苏大学电气信息工程学院,镇江212013

出　　处：《机械设计与制造》2012年第3期93-95,共3页Machinery Design & Manufacture

基　　金：江苏省自然科学基金资助项目(BK2009202)

摘　　要：实现虚拟轴机床末端刀具位姿的实时检测目前仍然是虚拟轴机床在数控加工领域实现高精度控制和产业化的障碍之一。针对六自由度虚拟轴机床的末端刀具位姿检测进行研究。首先对虚拟轴机床进行运动学分析,然后以虚拟轴机床末端刀具的位姿逆解作为神经网络的训练样本,构建结构自适应确定的RBF神经网络,实现虚拟轴机床从关节变量空间到工作变量空间的映射,最后利用已训练好的RBF神经网络实现虚拟轴机床末端刀具位姿的实时检测。实验结果表明:利用该方法实现虚拟轴机床末端刀具运动位姿的检测不仅具有可行性,而且具有较高的检测精度,为虚拟轴机床末端刀具的直接闭环高精度控制奠定了基础。Real-time pose measurement for terminal tool of a virtual axis machine tool is still one of obstacles to achieve high-precision control and industrialization in the field of digital control processing. Pose measurement of tool for 6-DOF virtual axis machine tool is addressed in this paper.Firstly,the kine- matics analysis of the virtual axis machine tool is made,then with pose inverse of terminal tool of virtual axis machine tool as neural network training samples ,a RBF neural network identified with self-adaptive structure is established,so that virtual axis machine tool realizes map from joint variable space to work vari- able space.Finally,the neural network,which have been trained,is applied to achieve the real-time pose measurement of its ＇tool.Experimental results show that ：the method is not only effective and high-precision to measure the position of tool of the virtual axis machine tool using RBF neural network identified with self-adaptive structure ,which lays the foundation for direct closed control of virtual axis machine tool.

关 键 词：虚拟轴机床 结构自适应 RBF神经网络 位姿检测 

分 类 号：TH16[机械工程—机械制造及自动化] TP216[自动化与计算机技术—检测技术与自动化装置]