摘要
本文利用ExtendedDelat－BarDelta（EDBD）算法的人工神经网络方法，对分光光度分析、流动注射分析及原子吸收3种典型体系中非线性定量关系进行了研究，对网络参数进行了优化，提出了适用于解决上述体系非线性定量关系的三种网络模型。结果表明，这些模型性能良好，预测精度高，为用人工种经网络解决扩大分析化学中的一些问题提供了一种新的手段，并使测试的应用范围扩大。
The nonlinear quantitative analysis of spectrophotometry, flow injection analysis and atomic absorption analysis were studied using artificial neural networks with extended delta-bar delta learning algorithm. The three optimized neural network models were created. The experimental results show that these models were good to solve the problems and it give us an alternative methods to expand the scope of the analysis methods.
出处
《分析化学》
SCIE
EI
CAS
CSCD
北大核心
1996年第6期724-728,共5页
Chinese Journal of Analytical Chemistry
关键词
神经网络
非线性定量
分光光度法
Artificial neural networks, spectrophotometry, flow injection analysis, atomic absorption spectrometry
