神经网络在高速高精度机器人测量反馈系统中的应用

讨论了神经网络在高速高精度机器人测量反馈系统中的应用。该系统采用了由激光发生器与精确的位置敏感检测元件组成的测量装置，该装置可测量挠性机械臂的除杆长方向外的五个自由度误差;在详细地论述了高速高精度机器人系统神经网络的学习与训练后，根据系统的不同要求分析了四种离线学习方案，并通过仿真计算加以证实。所介绍的方法为高速高精度机器人的实时控制提供了一种有效的途径。

Application of Neuro Network to the Feedback System in High speed High-precisionR obot Meassurement

This paper deals with the application of neuro network to the feedback system in high-speed high precision robot measurement. This system uses the measuring device consisting of the laser generator and the poition sensitive element. This device is capable to measure the errors of flexible arm in five freedom degree excepting in the longitudinal direction of the rod. After thoroughly describing learning and training of the neuro network of the high-speed high precision robot, this paper analyses four veriants of off-line learning according to different requirements which are verified through simulation caclulation This method presents an effective way to real time control the high speed high-precision robot.