摘要
本文报道化学计量学方法用于多环芳烃(PAHs)液相色谱分离条件的优化。使用均匀实验设计法,以乙腈在线性梯度展开时的初始浓度和线性梯度的斜率为优化参数,对16种多环芳烃混合体系进行液相色谱分离条件优化,采用遗传算法和退火神经网络方法建立了有效的分离条件预测模型。对模型所预测的最佳分离条件进行试验,分离结果满意。
The chemometrics methods were applied to the high performance liquid chromatographic（HPLC） separation of polycyclic aromatic hydrocarbons.A prediction model was established on the basis of the uniform test designs,and the starting concentration and the slope of CH3CN in linear gradient developing for separation of 16 polycyclic aromatic hydrocarbons were optimized.An effective separation model was constructed by genetic algorithms and annealing neural networks.Using the optimized parameters predicted by the model,satisfactory separation performance was obtained by the experiment.
出处
《分析科学学报》
CAS
CSCD
北大核心
2011年第4期475-478,共4页
Journal of Analytical Science
基金
南京大学生命分析化学教育部重点实验室开放基金(No.KLACLS08004)
关键词
退火神经网络
遗传算法
高效液相色谱
多环芳烃
梯度分离条件优化
Annealing neural networks
Genetic algorithms
High performance liquid chromatography
Polycyclic aromatic hydrocarbons
Optimization of gradient separation conditions
