人工神经网络在小麦条锈病预测中的应用
pplication of BP and RBF Neural Network in forecast of Wheat Stripe Rust
利用甘肃天水1976-2001年的病情气象资料，采用逐步回归方法分析了影响小麦条锈病流行的因子，并将其作为BP（back-propagation）和RBF（radial basis function）神经网络的输入，用1976-2000年的资料进行网络训练，2001年的数据作为测试.结果显示对于训练数据集，BP和RBF网络的平均相对误差分别为0.3875%和0.0026%；而对于测试数据集，BP和RBF网络的相对误差分别为7.9529%和3.5709%. 可见人工神经网络用于小麦条锈病的预测是可行的，并且逼近能力、分类能力和学习速度等方面均优于BP网络的径向基函数网络的预测效果更佳.
he objective of this paper was the development of radial basis function networks （RBF）for incidence of wheat stripe rust prediction. The RBF prediction results were compared to those obtained with the back-propagation networks (BP). Simulations showed that the BP provided an average relative prediction error of 0.3875% for the training set and 7.9529% for the test set. The RBF provided an average relative prediction error of 0.0026% and 3.5709% on the same data sets， respectively. The RBF， therefore， outperformed the BP in terms of the prediction accuracy. The result shows that the RBFN and BPN can be successfully applied to build the relation of the incidence of stripe rust and meteorological data.
马占鸿、刘荣英
植物保护生物科学现状、生物科学发展计算技术、计算机技术
小麦条锈病， BP ， RBF， 预测
Wheat Stripe Rust， BP， RBF， Forecast
马占鸿,刘荣英.人工神经网络在小麦条锈病预测中的应用[EB/OL].(2006-01-16)[2025-08-02].http://www.paper.edu.cn/releasepaper/content/200601-175.点此复制
评论