摘要
采用4_(2_吡啶偶氮)_间苯二酚(PAR) -Cu、Co、Ni显色体系 ,应用Madline神经网络原理 ,对于吸收光谱严重重叠的三组分金属配合物体系同时进行含量测定 ;Cu2 +、Co2 +、Ni2 +的平均回收率分别为94 %、101 %、102% ;实验表明 ,该法具有训练速度快、预测结果准确度良好等特点 ,有望成为多组分分析的有效方法之一。
Madaline neural network was used for the simultaneous determination of contents of Cu, Co and Ni using 4 _ (2 _ pyridylazo) _ resorcinol(PAR)-Cu, Co and Ni colour system in which the spectra of complexes of PAR with Cu, Co and Ni are seriously overlapped. The mean recoveries of Cu, Co and Ni are 94%, 101% and 102%, respectively. The network is better than other networks in training speed and accuracy, therefore, it is a prospective method for spectrophotometric analysis of multicomponent system.
出处
《分析测试学报》
CAS
CSCD
北大核心
2002年第2期64-66,共3页
Journal of Instrumental Analysis
基金
安徽省自然科学基金资助项目 (00046509)
