多源数据小麦条锈病预测研究

孔钰如,王李娟,张竞成,杨贵军,岳云,杨小冬

Research on Prediction of Wheat Stripe Rust with Multi-source Data

Yuru Kong,Lijuan Wang,Jingcheng Zhang,Guijun Yang,Yun Yue,Xiaodong Yang

表4 拟合模型

Table 4 Fit models

模型名称	方程式	参数
BPNN模型	-	隐藏层节点数为3
SVR模型	-	$\begin{array}{l} c = 0.35; \\ g = 0.5 \end{array}$
MLR模型	$Y = 0.21 \times N D V I + 0.01 \times T + 0.01 \times H - 0.35$	-
SEIR-StripeRust模型	$\begin{array}{l} \frac{d H}{d t} = - β H I \\ \frac{d L}{d t} = β H I - ω L \\ \frac{d I}{d t} = ω L - μ I \\ \frac{d R}{d t} = μ I \\ β = k_{a} \times T \times H + k_{b} + k_{c} \times N D V I \\ Y = L + I + R \end{array}$	$\begin{array}{l} k_{a} = 2; \\ k_{b} = 0.9; \\ k_{c} = 1; \\ 1 / ω = 18; \\ 1 / μ = 30 \end{array}$