摘要
应用人工神经网络原理，采用误差反向传播方法，对紫外吸收光谱严重重叠的复方替硝唑进行含量测定。实验表明，人工神经网络方法应用在光谱分析中对复方制剂进行含量测定，其结果准确，性能良好；并探讨了人工神经网络中网络参数对分析结果的影响。
By means of error backpropagation train algorithm, the principle of artificial neural network has been applied to determine the contents of Tinidazolum compound in which spectra of Tinidazohum and Nysfungin are overlapped. The calculation results show that determination of the medicine by spectroscopic analysis through artificial neural network is accurate. The effect of the network parameters on the results of the analysis has been discussed too.
出处
《中国药科大学学报》
CAS
CSCD
北大核心
1998年第2期128-131,共4页
Journal of China Pharmaceutical University
关键词
人工神经网络
替硝唑
含量测定
分光光度法
Artificial neural network
Backpropagation
Tinidazolum
Nysfungin
Determination of content
