基于BP神经网络的张家口市莜麦气象产量预测研究  

Research on Forecasting the Meteorological Yield of Naked Oats in Zhangjiakou City Based on BP Neural Network

作　　者：温欣茹 苌自力 叶永飞[1] 武贵贵 孙兴华[1] 李坤鹏 Wen Xinru;Chang Zili;Ye Yongfei;Wu Guigui;Sun Xinghua;Li Kunpeng(College of Information Science and Engineering,Hebei North University,Zhangjiakou,China)

机构地区：[1]河北北方学院信息科学与工程学院,河北张家口

出　　处：《科学技术创新》2023年第17期93-96,共4页Scientific and Technological Innovation

基　　金：大学生创新创业训练计划项目“基于气象数据的张家口市主要粮食作物产量预测研究”(202210092006);河北省高等学校人文社会科学研究项目“课程思政视域下信息工程专业‘双创教育’工程化实践模式研究”(SD2022048)。

摘　　要：利用河北省张家口市2002-2017年粮食生产相关数据,将特色作物莜麦作为主要研究对象,建立了基于气象因素的逐步回归分析预测模型和BP神经网络预测模型进行莜麦产量预测。模型中以苗期、分蘖期、灌浆-成熟期的日照时数,分蘖期、抽穗-开花期和灌浆-成熟期的降水量共6个指标作为主要的输入值,预测莜麦的气象产量。实验结果表明BP神经网络应用于莜麦产量预测精度高,对粮食产量预测具有指导意义,利于农事生产。Based on the grain production data of Zhangjiakou City in Hebei Province from 2002 to 2017,taking the characteristic crop naked oats as the main research object,a stepwise regression analysis prediction model and a BP neural network prediction model based on meteorological factors were established to predict the yield of naked oats.In the model,the Sunshine duration at seedling stage,tillering stage,filling maturity stage,precipitation at tillering stage,heading flowering stage and filling maturity stage were taken as the main input values to predict the meteorological yield of naked oats.The experimental results indicate that the application of BP neural network in predicting naked oats yield has high accuracy and guiding significance for predicting grain yield,which is beneficial for agricultural production.

关 键 词：粮食产量预测 莜麦 气象产量 BP神经网络 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]