主成分回归在水稻需水量预测中的应用

The Application of the Principal Component Regression in Forecasting Paddy Water Requirements

影响水稻需水量的主要因素有气温、水面蒸发、空气饱和差、日照和风速,这几个影响因子之间存在着相关性。运用多元统计分析中的主成分分析法通过提取两个主成分,克服变量之间的相关性,然后再进行回归,建立主成分回归的水稻需水量模型,并对模型进行了拟合质量检查,取得了较为满意的结果。

The main factors which impact water requirements of paddy are climate, air saturation difference, wind velocity, sunlight and evapotranspiration among which there is multi-collinearity. The principal component regression (PCR) is applied to set up the paddy water requirement model which can solve the problem of interactive correlation among the independent variables. The author has analyzed the PCA model and gained a satisfying result.