摘要
色谱图中两个(或以上)峰相互重叠时直接影响色谱分析的结果.本文提出一种借助于人工神经网络的解析方法,将色谱重叠峰切割成间隔为0.02分的“薄片”,以“薄片”面积为输入值,以前向多层人工神经网络处理.可以准确测定峰形重叠的二乙胺和三乙胺的含量,经标准物对比。
A new method for resolving chromatographic overlapping peaks with artificial neural network is proposed in this paper. The overlapping peak was divided into 15 piece(0.02 minute interval), and their areas were taken for the input data of an artificial neural network with back propagation algorithm. After comparing the peaks between standard and testing samples, the satisfactory result was obtained for determination of diethyl amine and triethyl amine.
出处
《杭州大学学报(自然科学版)》
CSCD
1998年第3期62-66,共5页
Journal of Hangzhou University Natural Science Edition
关键词
色谱分析
重叠峰解析
人工神经网络
chromatographic analysis
resolving of overlapping peak
artificial neural network