人工神经网络光度法用于化探样品中金的形态分析  

Species Analysis of Gold in Geochemical Samples by Artificial Neural Network

作　　者：王英华[1] 陈淑桂[1] 王洪艳[1] 欧阳立群[1] 樊玉清[2] 

机构地区：[1]吉林大学朝阳校区化学系,长春130026 [2]山东轻工业学院化学工程系,济南250100

出　　处：《分析化学》2002年第1期62-65,共4页Chinese Journal of Analytical Chemistry

基　　金：国家自然科学基金资助项目(No.29975004)

摘　　要：利用人工神经网络对化探样品中金的三种存在形态（水提取态、吸附态、有机态）的含量同时进行了预测，计算结果与传统浸取方法的结果相吻合，相对误差一般小于１０％，相对标准偏差在１．６８％～８，９４％。该法简便、快速。Artificial neural network (ANN) was applied to the simultaneous determination of gold species (water extractable species, adsorpted species and organic species) in geochemical samples. The relative deviations of the prediction to the reference values are less than +/-10%, the relative standard deviation is 1.7%similar to8.9% for the determination of geochemical samples (n=17). The calculated results of this method axe in agreement with those of traditional soaking method. The method is simple and quick.

关 键 词：形态分析 人工神经网络 金 化探样品 矿物 光度法 

分 类 号：O657.3[理学—分析化学]