人工神经网络用于高效液相色谱分析条件的优化

Study on Optimizing of Separation Conditions for Mixtures of Phenols With High-performance Liquid Chromatography by Using Artificial Neural Network

本文采用人工神经网络的线性网络法(LMS)优化高效液相色谱的分析条件,考察了样本集类型和不同初始权值对优化结果的影响.结果表明,模拟的精度与样本集类型有关,而在一定置信区间与初始权值无关.

One type of artificial neural networks, least means square, was adopted to optimize separation conditions for mixture of phenols with HPLC. The effect of type of samples and initial weight on prediction values was discussed. The experimental results showed that the precision of predicting values depends directly on the type of samples without relation to initial weight.