摘要
把人工神经网络用于六种氨基酸(酪氨酸、色氨酸、苯丙氨酸、胱氨酸、组氨酸,3,4—二羟基苯丙氨酸)混合物紫外光谱的定量分析,以酪氨酸为例,不经分离测定了混合溶液中酪氨酸的含量,为氨基酸的多组分分析提供了一种新方法。并与卡尔曼滤波方法进行了比较,表明神经网络在一些方面优于卡尔曼滤波方法。
The multicomponent quantitative analysis of amino acids is usually carried out ly high performance liquid chroma tography(HPLC) or special amino acids analyzer, but these methods are expensive and tedious. In this paper,we have successfully determined the amount of tyrosine in the six amino acids mixture (Tyrosine、Tryptophan、 Phenylalanine、 Cystine、 Histidine、3, 4-Dihydroxyl、 Phenylalanine) without any separation from UV spectra by neural network. It gives a new method for the quantitative analysis of multicomponent amino acids. Comparison with the method of Kalman filter algorithm is also given.
出处
《生物物理学报》
CAS
CSCD
北大核心
1992年第4期706-710,共5页
Acta Biophysica Sinica
关键词
神经元网络
紫外光谱
氨基酸
neural network, UV spectrophotometry, amino acids, multicomponent analysis