二次旋转设计和反向人工神经网络在铅、镉、汞、镍同时测定中的应用

APPLICATION OF THE GENERAL REVOLVING COMBINATION DESIGN AND BACKPROPAGATION- ANN IN THE SIMULTANEOUS SPECTROPHOTOMETRIC DETERMINATION OF LEAD, CADMIUM, MERCURY, NICKEL

利用二次通用旋转组合设计法建立了吸光度与显色剂、增敏剂用量及pH值之间的数学模型,进而得出使吸光度最大的显色剂用量、增敏剂用量及pH值。在优化的基础上,用B-P人工神经网络对Pb、Cd、Hg、Ni同时测定的数据进行解析,井与经典最小二乘法进行了对比,结果较准确。

A mathematical model has been established through the general revolving combination design to describe the absorbance responses as a function of sodium lauryl sulfate, 4-(2-pyridylazo)-resorcinol, pH values of the system. After optimum coefficients were found for the model, the determination of Pb, Cd, Hg and Ni in unknown were complete by using B-P ANN. The study on B-P ANN and PLS showed that the results were satisfactory.