序贯实验设计法在吸附平衡实验中的应用

采用序贯实验设计法对正葵烷在5A分子筛上的吸附等温线实验进行了实验设计。根据Box和Lucas所提出的使待估参数联合置信域的容积趋于最小的参数估算的序贯实验设计准则，推导得出了正葵烷吸附等温线实验观测点的安排原则，发现对于一元线性模型，实验观测点应安排在可操作区的两端。对序贯设计实验和未经设计实验做了比较，经过序贯设计的实验所得到的模型参数的联合置信域比未经设计实验的小得多，表明序贯实验设计可提高模型参数估算的精度，减少实验工作量。

Adsorption isotherm parameter estimation by using sequential experimental design technique

A sequential experimental design procedure was applied to optimize parameter estimation for adsorption isotherm of n -decane on 5A molecular sieves. Based on the criterion that the volume of estimated parameter's joint confidence region should be minimum, which is a sequential experimental design norm for optimal parameter estimation proposed by Box and Lucas, the observations for the adsorption of n -decane on 5 A molecular sieves were determined sequentially, and it is found that the observations should be located on the both sides of the operational region when mathematical model is mono linear, and the volume of estimated parameter's joint confidence region emerging from sequential designed experiments was much smaller in comparison with experiments without design. It shows that sequential experimental design can improve accuracy of parameter estimation and reduce laboratory work.