径向基函数网络光度法用于多组分分析  

Radial Basis Network Spectrophotometry Applied to Analysis of Multicomponent

作　　者：王建[1] 刘建华[1] 潘艳玲[1] 陈国南[1] 

机构地区：[1]福州大学化学系,福州350002

出　　处：《计算机与应用化学》2000年第3期262-266,共5页Computers and Applied Chemistry

摘　　要：对神经网络在多组分光度分析法中的应用进行了研究。应用径向基函数网络和BP网络 ,采用紫外光度法同时分析了色氨酸、酪氨酸和苯丙氨酸三元组分 ;采用速差动力学光度法同时分析了铜、锌、铁混合组分。实验证明 ,径向基函数网络在所需神经元个数、训练时间、预测准确度等方面均明显优于BP网络算法。该技术和光度法结合有望成为多组分分析的有效选择方法之一。The application of neural network in the multicomponent analysis is presented. The multicomponent of amino acids have been determined with the radial basis network and BP network UV spectrophotometry, the Cu(Ⅰ),Zn(Ⅱ) and Fe(Ⅲ) complex have been determined with the radial basis network and BP network differential reaction rate kinetic spectrophotometry.The results obtained with radial basis network are much better than those provided by BP network in training time,the number of neurons and the accuracy. In conclusion,the radial basis network spectrophotometry is a good choice for resolving multicomponent.

关 键 词：径向基函数网络 速差动力学光度法 多组分分析 

分 类 号：O657.3[理学—分析化学]