摘要
良好的试验设计方案在农业优化栽培技术中具有重要意义,本文分析了常用的二次旋转回归设计的预测值误差特性,找出了此设计的合理预测区间,根据农业优化栽培试验的特点,提出了一种新的优化栽培试验设计方案,该设计在各因子取值区域边沿的预测值方差较原设计可降低90％以上。
Analysing the error characteristic of values calculated by ordinary second order rotatable regression, the authors found out the reasonable calculation range of design and proposed a new design programme for optimum cropping experiments by which the variance of calculated values in the edge of each factor range can decrease 6—11 times than that by ordinary designs.
出处
《山东农业科学》
1992年第1期1-7,共7页
Shandong Agricultural Sciences
关键词
栽培
回归设计
最佳化
optimum, agricultural cropping, regressive design
