人工神经网络光度法同时测定水中Fe(II)、Fe(III)

Simultaneous Spectrophotometric Determination for Fe(II)、Fe(III)by Artificial Neural Networks

应用径向基函数人工神经网络(RBF)和误差反向传播神经网络(BP),采用磺基水杨酸-邻菲罗啉双显色剂光度法同时分析了水中Fe(II)和Fe(III)。实验证明,RBF算法在所需神经元个数、训练时间、预测准确度等方面均明显优于BP网络算法。该技术和光度法结合有望成为多组分分析的有效选择方法之一。

The application of radical basis network （RBF） and BP neutral Network （BP）to the simultaneous determination of Fe （Ⅱ）and Fe （Ⅲ） has been accounted. Mixed reagents of sulfo-salicylic acid and phenenthroline was used as a selective chromogenic system. It was proved that RBF is prior to BP comparing to the number of nerve units, training time and pre-accuracy. It will be one of best selective methods to determine multiple components when combining RBF with photometric method.