基于灰色系统理论的四川油料作物单产预测

为准确预测四川油料作物的产量以保障市场供给,根据2001—2015年四川油料作物单产数据,运用灰色系统理论,建立GM(1,1)预测模型,并通过残差、级比偏差、关联度、后验差检测及模拟数据检查对模型的合理性和精度进行误差检验,并应用模型预测未来3年四川油料作物的单产。结果表明:P~0=98.22%90%、ρ(avg)=4.52%10%、r=0.97760.6、P=1,F=1.1141F_(0.05)=19.00,由此可见,灰色系统理论GM(1,1)适用于油料作物单产预测,并具有较高精度。预测得出,2016年、2017年、2018年的油料作物单产分别为162.25kg/667m~2、164.71kg/667m~2和167.22kg/667m~2,与2015年相比,分别上涨2.70%、4.26%和5.84%。

Unit Yield of Oil Crops Predicted by Grey System Theory in Sichuan

The GM(1,1) prediction model is built by Grey System Theory according to unit yield data of oil crops in Sichuan from 2001 to 2015,the rationality and accuracy of the built GM(1,1) prediction model are tested by residual error,grade ratio deviation,correlation degree,posterior difference test and simulated data and the unit yield of oil crops in 2016,2017 and 2018 is predicted by the built GM(1,1)model to forecast yield of oil crops for safeguarding market supply in Sichuan.Results:The error test results are P0=98.22％＞90％,p(avg)=4.52％＜10％,r=0.9776＞0.6,P=1 and F=1.1141＜F0.06 =19.00 respectively,which indicates the built GM (1,1) model with a high accuracy can be used in prediction of oil crops' unit yield.The unit yield of oil crops in 2016,2017 and 2018 predicted by the built GM(1,1) model is 162.25 kg/667m2,164.71 kg/667m2and 167.22 kg/667m2,2.70％,4.26％ and 5.84％ higher than the unit yield in 2015 separately.