论文部分内容阅读
根据作物产量与环境因子之间的实际情况，以东港市水稻产量资料和气象资料为例，采用非线性的最优化方法普查预报因子，通过因子的显著性检验和稳定性检验，选择出最优预报因子。然后根据农业气候分析的方法，从最优预报因子中选择出重要因子，用强迫引进重要因子和逐步回归方法，建立最优的非线性预报方程。通过验证，不但拟合率高，而且试报效果也非常好。同时本方法可应用于农业气象的其它方面的最优化问题的预测预报中。还可推广到预报量离散的情况，建立非线性的判别方程。
According to the actual situation of crop yield and environmental factors, taking the rice yield data and meteorological data of Donggang as an example, we use the non-linear optimization method to censor the forecasting factors and choose the optimal one through the significance test and stability test Predictors. Then, according to the method of agroclimatic analysis, choose the important factor from the optimal forecasting factor, and establish the optimal nonlinear forecasting equation by forced introduction of important factors and stepwise regression. Through the verification, not only the fitting rate is high, but also the test results are very good. At the same time, the method can be applied to the prediction and forecast of other optimization problems of agrometeorology. Can also be extended to predict the amount of discrete cases, the establishment of nonlinear discriminant equations.