基于决策树算法的安徽省油菜产量气象限制因子分析及预测模型研究

Analysis and prediction model of meteorological restriction factors of Anhui province rapeseed single yield based on decision tree algorithm

为分析气象因子对安徽省油菜(Brassica napus L.)产量的影响,构建适用于油菜单位面积产量预测模型,利用1999—2018年安徽省78个站点的地面气象观测资料及2000—2018年78个县(市、区)的油菜单位面积产量数据,采用决策树算法对影响油菜产量的气象因子进行分析,筛选因子基于支持向量机建立油菜产量预测模型。结果表明,影响油菜产量的主要气象因子是成熟期湿润指数、蕾薹期平均气温、苗期湿润指数、开花期湿润指数,预测模型的预测值与实测值的均方根误差为402 kg/hm2,拟合指数为0.72。

Aiming to construct a local rapeseed single yield forecasting model based on analysis of the influence of meteorological fac⁃tors on rapeseed yield in Anhui region,the datas containing surface meteorological data of 78 observation sites in Anhui province from 1999 to 2018 and rapeseed single yield of 78 counties(cities,districts)from 2000 to 2018 were collected.The decision tree algorithm was adopted to analyze influence of meteorological factors on rapeseed output which finally screened out the main factors.A predictive model of rapeseed output based on support vector machine(SVM)was constructed.The results showed that the main meteorological factors which affected rapeseed yield were wetting index during maturity,average temperature during bud stage,wetting index at seed⁃ling stage and flowering stage.The root mean square error between predicted and measured value was 402 kg/hm2,and fitting index was 0.72 which proved the prediction model behaves well on the whole.