用神经网络和分光光度法测定废水中的多元素  

Determination of Multielement in Industrial Wastewater Using Artifical Neural Network and Spectrophotometry

作　　者：管棣[1] 谢青兰[1] 王纪明[1] 张媛媛[1] 

机构地区：[1]西南交通大学生物工程学院,四川成都610031

出　　处：《西南交通大学学报》2007年第1期129-132,共4页Journal of Southwest Jiaotong University

基　　金：四川省科学技术厅应用基础研究基金资助项目(04JY029-001-5)

摘　　要：为了快速测定工业废水中吸收光谱重叠的Cd与Co和Cr与Fe 2组元素,采用4-(2-吡啶偶氮)-间苯二酚(PAR)-Cd-Co和二苯碳酰二肼-Cr-Fe显色体系,采用分光光度法测定吸光度,用训练过的人工神经网络进行数据处理,建立了不经分离而能同时测定Cd与Co,Cr与Fe含量的方法.用此方法测定合成样品,Cd,Co,Cr和Fe的回收率分别为101%,98%,102%和98%,平均相对误差分别为1.61%,-0.23%,0.00%和-0.08%.In order to quickly test two groups of elements with overlapping absorption spectrum, Cd and Co, and Cr and Fe, in industrial wastewater, a novel method was proposed. With this method, the contents of the two elements in each of the two grounds can be tested synchronously and they need not to be separated. In the method, the color system of4-（2-pyridyl azo）-resorcin （PAR） -Cd-Co and diphenylcarbazide-Cr-Fe is adopted, the spectrophotometry is utilized to test absorbency, and a drained artificial neural network is used to deal with data. The tested results of synthetical samples show that the recovery ratio of Cd, Co, Cr and Fe is respectively 101% , 98% , 102% and 98% , and the mean relative error is 1.61% , -0.23% , 0.00% and -0.08% respectively.

关 键 词：人工神经网络 分光光度法 工业废水 CD Co CR Fe 

分 类 号：X832[环境科学与工程—环境工程]