人工神经网络在三化螟预测中的应用  

The application of neural networks to Scirpophaga incertulas forecasting

作　　者：林高飞[1] 林乃铨[2] 尤民生[3] 

机构地区：[1]福建农业大学机电工程系 [2]福州生物防治研究所 [3]福州植物保护系,福州350002

出　　处：《华东昆虫学报》1999年第2期74-77,共4页Entomological Journal of East China

摘　　要：本文应用神经网络对闽北四代三化螟的发生情况进行了预测。三化螟的预报可视作高维空间的非线性分类问题。利用ＢＰ网络的反向传播算法，我们对三化螟种群发生趋势进行预测，并获得了满意的结果。This paper presents the application of neural networks to the prediction of the fourth generation of Scirpophaga incertulas (Walker) in northern area of Fujian province.The prediction of the occurrence of Scirpophaga incertulas (Walker) can be regarded as the nonlinear classification problem in the highdimension space.Using the back- propagation algorithm,we have predicted the occurrence of the pest and given relatively satistactory result.

关 键 词：三化螟 发生趋势 闽北 发生情况 种群 利用 应用 分类问题 高维空间 

分 类 号：Q959.52[生物学—动物学] S435[农业科学—农业昆虫与害虫防治]