人工神经网络方法用于多组分分光光度分析

Spectrophotometric Analysis of Multicomponent-multisystems Using Artificial Neural Network

本文介绍了人工神经网络(ANN)原理,详细讨论了网络参数的选择及其对网络预报结果的影响。并将人工神经网络应用于干扰严重的五组分体系的分光光度同时分析,其预测结果优于正交分解计算方法。

This paper deals with the principles of artificial neural network. The effects of various ANN parameters on prediction are discussed. The ANN method has been applied to the simultaneous spectrophotometrlc analysis of multlcomponent-multisystems. The results show that the ANN method is better than the orthogonal decomposition method in the spectrophotometric analysis of muhicomponent-multisystems.