摘要
本文根据二次正交旋转组合设计的特点和局部控制理论,提出了中心点数据处理法。此法是先求出各区组的矫正系数,即各区组中心点总平均值与本区组中心点平均值之比,然后将小区原始数据分别乘本区组的矫正系数。用此法排除非试验因子干扰的效果明显地优于飘移处理法,不仅计算简单,而且有拟合程度高、便于模拟选优的优点。
According to the partial-control theory and the traits of design of quadratic rotation-regression-orthogonal combination, the central point treatment is put forward in this paper. In Consideration of the elimination of interference caused, by non-test factor the central point treatment is obviously superior to the drift treatment. In addition, this method is easy to use.
关键词
正交
旋转组合设计
中心处理法
Quadratic rotation-regression-orthogonal combination, Central point treatment
Drift treatment.
