本文应用多元二次回归正交旋转组合分段设计方法,研究了九因素对小麦产量的综合影响。中间选用冬前苗情做关键指标,承上启下使两段试验联成一体。为摸清前段措施后效作用对产量的影响,通过反馈计算求得后效产量Δy_1进而按试验设计配置x_6(根据前段苗情)、y_2(根据后段产量),Δy(根据前段实际产量与理论产量之差)三个多元二次回归数学模型,并将三者组合运用,进行综合寻优或分段决策,收到了良好效果。 In this paper, the multiple quadratic regression orthogonal rotation combined subsection design method was used to study the combined effects of nine factors on wheat yield. The middle selection of winter seedlings do the key indicators, bearing on to make the two tests into one. In order to find out the effect of the aftereffect of the preceding measures on the yield, the post-production yield Δy_1 was obtained by the feedback calculation, and then the experimental design was used to configure x_6 (according to the previous stage seedling), y_2 (according to the latter stage yield), Δy The difference between the theoretical yield) three multiple regression model of multiple regression, and the combination of the three, the comprehensive optimization or segmentation decision-making, received good results.