南方地区冬小麦白粉病预测模型的研究  

Study on Prediction Model for Wheat Powdery Mildew in South Area of China

作　　者：李光泉[1,2] 陈琦[1] 胡亚平[1] 何火娇 

机构地区：[1]江西农业大学计算机与信息工程学院,南昌330045 [2]江西省高等学校农业信息技术重点试验室,南昌330045

出　　处：《农机化研究》2011年第5期49-51,55,共4页Journal of Agricultural Mechanization Research

基　　金：江西省教育厅科技项目(赣教技字[2006]148号)

摘　　要：由于南方地区冬小麦白粉病的预测是一项多因素的复杂工程,因而针对传统预测方法的不足,提出了一种基于灰色系统GM(1,1)理论和BP神经网络相结合的预测模型。灰色系统GM(1,1)模型理论建模方法简单,计算量小,而BP算法特别适合于对GM(1,1)模型进行残差修正,利用两者的结合可以建立较高精度的预测模型。利用该模型对南方某地区冬小麦白粉病进行了预测,结果表明,该模型准确有效,适用于该地区的冬小麦白粉病的预测。The prediction of winter wheat powdery mildew in the south area is a multi-factor complex projects,as the deficiencies of traditional forecasting methods,a new prediction model which combined the grey system GM（1,1） theory with the BP neural network is proposed;because grey GM（1,1） model theory is simple and computation,and BP algorithm is especially suitable for GM（1,1） model residual modification,it can be established using a combination of high precision prediction model;the winter wheat powdery mildew of one region in south area was predicted by using the model,results showed that the model is accurate and effective for the region forecast of winter wheat powdery mildew.

关 键 词：冬小麦白粉病 预测模型 灰色系统 神经网络 

分 类 号：S435.12[农业科学—农业昆虫与害虫防治]