基于BP神经网络的二维精密扫描台控制技术及试验研究  

The Control Technology and Experimental Study on 2-D Precise Scanning Stage Based on BP Neural Network Method

作　　者：刘梅[1] 李玉和[1] 郑静琳[1] 林浩山[1] 

机构地区：[1]清华大学精密仪器与机械学系,北京100084

出　　处：《新技术新工艺》2008年第11期64-66,共3页New Technology & New Process

基　　金：973国家重大基础研究项目(2003CB716201)

摘　　要：扫描控制系统是纳米测量系统的重要组成部分,其性能影响整个系统的测量精度。本文介绍了二维精密扫描台的设计——柔性铰链与压电控制结合的方式,综述了BP神经网络的概念,并通过BP神经网络建模,设计了扫描台的控制方法,最后介绍了试验设备,并通过试验测试得到了最佳控制参数,验证得到建模控制方案与实际位移相吻合。The scanning system is one of the most important parts of the nano-measuring system,its capability influences on the main performance of the system.This paper introduces the structure of a 2-dimensions precise scanning stage that means the method mixed flexible hinge and piezoelectric control.It also gives a summary of BP neural network concepts,and the princi- ple of BP neural network method.The paper makes a modeling using BP neural network method to realize the control on scan- ner.The experiment presented in the end shows that the model- ing matches the actual results perfectly.

关 键 词：压电陶瓷 BP网络 神经元 隐层 

分 类 号：TP273.22[自动化与计算机技术—检测技术与自动化装置]