基于BP神经网络紫外可见光度法测定苏丹红混合组分

应用人工神经网络,以BP算法对混合样品中苏丹红系列的三种组分(苏丹红Ⅰ、苏丹红Ⅲ、苏丹红Ⅳ)的浓度进行测定。在MATLAB 7.0中建立BP神经网络,优化网络条件,对训练样本进行训练,然后对检测样本进行检测。预测结果的误差范围在0.03%~9.20%之间。当样品浓度<0.1×10-4mol/L时,预测误差较大,均在5%以上;当浓度>0.1×10-4mol/L时,预测的相对误差较小,均在5%以下。该方法已用于模拟水样中微量苏丹红的检测。

Application of BP neural network and UV-Visible spectrophotometry to predict mixed components of Sudan

Mixed samples of the three components series Sudan (Sudan Ⅰ,Sudan Ⅲ,Sudan Ⅳ) concentrations were measured applying artificial neural networks to BP algorithm.BP artificial neural network model was constructed by MATLAB 7.0,and network conditions were optimized.Simulated data were used to train the net work so as to gain the suitable learning epochs and the size of learning set.The effects of neuron number in hidden layer and momentum parameter on classification were investigated.All the predicted values by the model have been compared with the practical values and the errors were between 0.03% and 9.20%.When the concentration of sample is less than 0.1×10~(-4) mol/L,the prediction error is more than 5%;when the concentration is greater than 0.1×10~(-4) mol/L,the relative error is below 5%.The method using BP artificial neural network model to predict the series of Sudan is feasible.