应用人工神经网络优化分离抗菌药物混合物的液相条件

Optimization of HPLC Gradient Separation of Mixed Antibacterial Agents by Artificial Neural Networks

将人工神经网络法应用于抗菌药物高效液相分离条件的优化。采用正交试验法,以流动相中乙腈初始浓度、线性梯度斜率及pH为优化参数,对7种抗菌药物混合体系进行优化。采用误差反向传输方法建立了神经网络权接拓扑模型,预测最佳分离条件,获得了满意的分离结果。

The artificial neural networks （ANN） was applied to the gradient separation of 7 antibacterial agents in mixture by HPLC. A prediction ANN model was established on the basis of the orthogonal design, and the parameters such as starting concentration of acetonitrile, the gradient slope and pH value of the buffer solution in linear gradient developing were optimized. Satisfactory separation was obtained with the optimized parameters predicted by the proposed back propagation ANN model.