RBF神经网络在红外CO_2传感器压力补偿中的应用研究  

Study on application of RBF neural network in pressure compensation of IR carbon dioxide sensor

作　　者：钱力[1] 傅岚[1] 黄刚[1] 

机构地区：[1]中国航天员科研训练中心,北京100094

出　　处：《传感器与微系统》2008年第1期30-32,共3页Transducer and Microsystem Technologies

摘　　要：在红外CO2传感器的测量过程中,环境总压是一个重要的影响因素。在环境总压变化的情况下做好压力补偿得出正确的CO2气体分压值,对提高传感器的测量精度有重要意义。提出一种基于聚类和梯度法的径向基函数(RBF)神经网络方法,利用它的局部逼近特性,建立起其在红外CO2传感器的非线性压力补偿中的网络模型。实验结果表明:该应用收到了良好的效果。Environment pressure is an important effective factor in measuring process of IR carbon dioxide sensor. Doing pressure compensation well to get the accurate carbon dioxide gas pressure data when environment pressure changes,which makes great sense for the improvement of measurement accuracy of the sensor. Here a method of radial basis function （ RBF ） neural network based on clustering and gradient is presented. By using its part approaching ability,the net model is established in the non-linearity pressure compensation. The experiment results show that the application gets well effect.

关 键 词：红外CO2传感器 径向基函数神经网络 压力补偿 

分 类 号：TP212[自动化与计算机技术—检测技术与自动化装置] TP183[自动化与计算机技术—控制科学与工程]