基于径向基神经网络的水饱和含氧量的数据拟合  

Data-Fitting Calibration for Oxygen Saturation of Water Based on Redial Basis Neural Network

作　　者：琚新刚[1] 董乐 

机构地区：[1]河南教育学院电路与系统重点学科组,河南郑州450046 [2]郑州文理学院教务处,河南郑州450052

出　　处：《河南教育学院学报（自然科学版）》2011年第3期32-34,共3页Journal of Henan Institute of Education(Natural Science Edition)

基　　金：国家科技支撑计划(2006BAK01A38);河南省教育厅科技攻关项目(2009A510003);郑州市科技攻关项目(10PTGG379-1)

摘　　要：相关部门提供的标准大气压下水饱和含氧量标定值仅为离散温度点处的数据,这些数据显示了水的饱和含氧量与温度之间呈非线性关系,由此提出了利用Matlab函数创建径向基神经网络对既有标定数据进行分析、拟合.结果显示,该方法达到了误差目标,且较传统方法具有数据存储量小,网络学习时间短,收敛速度快的特点.Calibration data in the discrete temperature, provided by relevant departments, show a non-linear relationship between oxygen saturation of water and temperature. Meanwhile, calibration data are analyzed and fitted in radial basis neural network, which is constructed by a radial basis function in Matlab environment. The fitting resuit, obtained by the neural network simulation, shows that the method has reached the error goal, the amount of data storage is small, learning time is short, and convergence speed is fast.

关 键 词：径向基函数 神经网络 拟合 

分 类 号：TP391.9[自动化与计算机技术—计算机应用技术]