摘要
科学合理地考虑各生产资源要素在农业中的运用,促进粮食产量的稳定增长,是保障粮食安全的关键。选取河南省2005~2021年粮食产量及相关因素数据,利用灰色关联模型提取影响粮食产量的主要影响因素,基于方差倒数加权法构建由GM(1, N)、Lasso回归、BP神经网络组成的多变量变权重组合预测模型,对河南粮食产量的变化趋势进行拟合与预测。结果表明,变权重组合预测模型的预测误差为0.589%,预测精度高且性能稳定;预测河南粮食产量在2022~2025年将会保持稳定增长,并在2025年达到73 282.65 kt。
It is the key to ensure food security to scientifically and reasonably consider the application of various production resource elements in agriculture and promote the steady growth of grain output. The data of grain production and related factors in Henan Province from 2005~2021 were selected. And grey correlation model was used to extract the main influencing factors of grain production. Based on the inverse variance weighting method, a multivariate quantitative weight combination prediction model consisting of GM(1,N), Lasso regression and BP neural network was constructed. The change trend of grain output in Henan Province was fitted and predicted. The results showed that the prediction error of the variable weight combination prediction model was 0.589%, which had high prediction accuracy and stable performance. It was predicted that grain production in Henan would maintain a steady growth in 2022~2025 and reach 73 282.65 kt in 2025.
作者
张雅玉
李佳欣
王丰效
ZHANG Yayu;LI Jiaxin;WANG Fengxiao(School of Mathematics and Statistics,Kashi University,Kashi 844000,Xinjiang,China)
出处
《农业装备与车辆工程》
2024年第6期155-160,共6页
Agricultural Equipment & Vehicle Engineering
关键词
粮食产量
灰色关联分析
方差倒数加权法
变权组合预测
grain output
grey correlation analysis
inverse variance weighting method
variable weight combination prediction
