论文部分内容阅读
阐述了主分量分析法的原理与步骤,分析了河南省中原地区1990～2007年小麦白粉病病情及相关气象资料,得出影响其流行的主要分量,最后利用得到的主要分量作为BP神经网络的输入,对中原地区2008～2010年小麦白粉病流行情况进行预测,并与未进行主分量分析而建立的全要素BP网络模型进行比较。结果表明,该模型可以快速准确地预测小麦白粉病的流行程度,有效地减少小麦产量损失。
The principles and steps of principal component analysis (PCA) are expounded. The disease status and related meteorological data of wheat powdery mildew from 1990 to 2007 in Central Plains of Henan Province are analyzed. The main components that affect the prevalence of wheat powdery mildew are obtained. Finally, the main components obtained are used as BP neural network Inputs were used to predict the prevalence of powdery mildew in wheat from 2008 to 2010 in the Central Plains and compared with the total factor BP network model that was not established for principal component analysis. The results show that the model can predict the prevalence of wheat powdery mildew quickly and accurately, and effectively reduce the yield loss of wheat.