基于灰色数列预测模型的耕地变化预测——以贵州省普安县为例

Prediction of Cultivated Land Change Based on Gray Series Forecasting Model

简述了贵州省普安县气候及耕地面积情况,介绍了构建灰色数列预测模型的6个步骤,即序列生成、均值生成、建立GM(1,1)模型、还原生成、可靠性检验、外推预测。根据1998～2007年普安县的耕地面积数据,运用灰色系统理论建立了普安县耕地面积灰色数列预测模型,该模型通过了可靠性检验,拟合精度较高,可进行外推预测。基于构建的灰色数列预测模型对普安县2009～2012年的耕地面积变化情况进行了外推预测。预测结果表明,2009～2012年,普安县预期耕地面积将减少1 050 hm2,年均减少270 hm2,耕地面积呈逐年下降的趋势;预测模型结果与实际值接近,预测误差较小,是基本可信的。

Climate and cultivated land area of Puan County,Guizhou Province,China are briefly described.The six steps of Gray Series Forecasting Model are introduced,including series generation,mean generation,establishing GM（1,1） model,reducing reaction,reliability testing,and extrapolation forecast.According to the data of cultivated land area in Puan County from 1998 to 2007,Gray System Theory is used to establish the Gray Series Forecasting Model of cultivated land area in Puan City.Since it has passed the reliability test,this model has relatively high fitting precision and can be used for extrapolation forecast.Based on this Gray Series Forecasting Model,extrapolation forecast of cultivated land change is conducted in Puan County from 2009 to 2012.Result shows that the expected area of cultivated land will be reduced by 1050 hectares from 2009 to 2012,an annual decrease of 270 hectares.Cultivated land area shows a declining trend year by year.Result of the forecasting model is close to actual value with small forecast error.Thus,it can be concluded that the result is basically credible.