大豆产量预测预控数学模型的研究

STUDY ON THE MATHEMATICAL MODEL FOR FORECASTING AND CONTROLLING OF SOYBEAN

本试验采用四元二次正交回归旋转组合设计方法,建立了密度、氮、磷,钾肥四因素的综合效应产量模型,通过计算机模拟,筛选出亩产225公斤以上的农艺措施组合方案,并通过主成分分析,评价了各因素对产量的贡献。还利用特定产区大豆的历史资料,经逐步回归分析建立了相应产区的大豆气象趋势产量模型,并将气象因素引入栽培模型,建立了具体产地的大豆产量预测预控模型。

The synthetic efficiency model which concers four factors (X1: density; X2:nitrogen;X3: phosphorus; X4: patassium)is established with rotational design of orthogonal regression of quar-ternary quadratic form. The agronomical synthetic plan which the land can produce more than 225 kg/mu was selected, and the efficiency of every single factor and their interaction for every agronomical practice were analysed by wing the simulation of computor. The order of the four factors that contribute to the yield is: X1>X2>X3>X4.The models of meteorogical trend of the yield are also established for a special region bymeans of regressive analysis of the historic data (annual average yield from 1961--1987 andmonthly average tempratu re, sunshine hours and rainfall from may to September during the same time. ) in same regions. We also explorated the method of introducing meteorological factors into the cultural model of forecasting and contronlling for soybean in some special regions.