径向基神经网络补偿火焰光度氯传感器非线性输出的研究  

Study on nonlinear error compensation for FPCD based on RBF

作　　者：张道山[1,2] 张朝晖[1] 李志军[2] 李广勤[2] 赵亚彬[2] 王普红[2] 

机构地区：[1]北京科技大学信息工程学院,北京100083 [2]防化研究院,北京102205

出　　处：《现代仪器》2011年第2期60-62,共3页Modern Instruments

基　　金：国家自然科学基金(60977065)项目的资助

摘　　要：火焰光度氯传感器在气体检测上是一种很重要的方法。要提高其性能,它的非线性误差必须补偿。补偿的环节可以通过RBF的方法采设计,本文利用Matlab工具设计这个网络,实现非线性特性无失真地同实际物理过程一致的线性化,并减小非线性误差。结果表明,在FPCD设备上利用RBF补偿非线性是一个有效方法。FPCD is an important method in gas measurement .If the characteristics of the FPCD would be enhanced .The nonlinear errors had to be compensated .The compensation element was desired by the method of the RBF .This paper uses Matlab tools to design and train the net ,and the nonlinear characteristic of FPCD is changed into linear characteristic without distortion ,and identical to real physical process ,and thus nonlinear error is decreased .The result shows that this method is efficient via RBF application in the FPCD apparatus

关 键 词：非线性补偿 神经网络 火焰光度氯传感器 

分 类 号：TP183[自动化与计算机技术—控制理论与控制工程]