摘要
以安徽省农业总产值为应用背景,选择国家统计局的数据作为基础数据,分别采用GM、DEGM、RGM和DERGM四种模型对2011年至2017年安徽省农业总产值进行预测,并与实际总产值进行比较。得出四种模型的平均绝对百分误差(MAPE)分别为1.8257%、1.8023%、1.5367%和0.9252%,数值结果显示RGM的预测精度比GM和DEGM的精度都要高,说明滚动机制的应用能够提高GM(1,1)模型的预测精度;DERGM模型的预测精度是最高的,其MAPE的值为0.9252%,说明同时融入差分进化算法(DE)和滚动机制的GM(1,1)模型的性能得到大幅度提高。采用DERGM对2018—2020年安徽省农业总产值和全国农业总产值进行预测,得出2020年安徽省农业总产值将达到2 611.6亿元,所占全国农业总产值的比重约为3.80%。
Taking total agricultural output value of Anhui Province as application background and data of State Statistics Breau as essential data, this study adopts the GM, DEGM, RGM and DERGM respectively to predict the total agricultural output value of Anhui Province from 2011 to 2017, and compares the results with actual total output value. It is found that: mean absolute percentage errors (MAPE) of the four models arc 1.8257%, 1.8023%, 1.5367% and 0.9252% ; the RGM's predicting accuracy is higher than that of the GM and DEGM, which indicates that application of rolling mechanism can improve the GM (1, 1)'s predicting accuracy; the DERGM's predicting accuracy is the highest, of which the MAPE is 0.9252%, which indicates that integration of differential evolution (DE) algorithm and rolling mechanism greatly improves the GM (1, 1); predicted with the DERGM, the total agricultural output value of Anhui Province in 2020 will reach to 261.16 billion yuan, accounting for 3.80% of the total agricultural output value of China.
作者
王苗苗
罗清文
王瑞
WANG Miaomiao, LUO Qingwen, WANG Rui(School of Economics& Management, Anhui Agricultural University, Hefei 230036, China)
出处
《安徽农业大学学报(社会科学版)》
2018年第6期50-55,共6页
Journal of Anhui Agricultural University:SOC.SCI.
基金
国家自然科学基金项目“动态柔性碳配额下企业产品设计与产品定价策略研究”(71802004)
安徽省自然科学基金项目“基于事前决策的两阶段以旧换新运营策略研究”(1808085QG231)
关键词
灰色预测
滚动机制
差分进化算法
农业总产值
grey prediction
rolling mechanism
differential evolution algorithm
total agricultural output value