摘要
应用灰色系统理论和最小二乘原理, 推导出双变量连续微分 C G M (1 , 1 , 2) A 模型, 并将该模型用于高效液相色谱中溶质保留行为的预测。首次用建立的 C G M (1 , 1 , 2) 模型研究了不同流动相、不同色谱柱下几组烷基苯和链烷醇的容量因子与分子碳数和柱温的关系, 预测值平均相对偏差不超过1 . 5 % 。结果表明, 该模型适合描述溶质保留行为与分子结构参数之间的复杂关系, 为分子结构- 色谱保留定量相关 ( Q S R R) 研究提供了一种新的有效方法。
A continuous differential double variable gray model GM(1,1,2)A is derived based on gray system theory and the least-square method to predict the capacity factors of n\|alkylbenzenes and n\|alkanols by chain carbon number and column temperatures, which for the first time overcomes the deficiency of the unitary variable gray models. The relative deviations of the prediction are no more than 1.5%, as prove that GM(1,1,2) is completely adaptable to the studying on the retention behavior in HPLC. The results show it is a new effective method for Quantitative Structure Retention Relationship studying.
出处
《计算机与应用化学》
CAS
CSCD
1999年第5期385-386,共2页
Computers and Applied Chemistry
基金
国家自然科学基金