用专业统计资料进行稻麦病虫超长期预测的研究  

STUDY ON SUPER LONG TERM PREDICTION OF DISEASES AND INSECT PESTS OF RICE AND WHEAT WITH THE PROFESSIONAL STATISTICAL DATA IN SICHUAN

作　　者：刘牛[1] 李媛[1] 廖华明[1] 

机构地区：[1]四川省植保站

出　　处：《植保技术与推广》1998年第2期5-8,共4页Plant Protection Technology and Extension

摘　　要：本文以四川省植保专业统计资料为基础数据，提出了水稻、小麦主要病虫发生程度的分级标准。采用周期分析、改进特尔菲（Ｄｅｌｐｈｉ）、积分自回归滑动平均（ＡＲＨＭＡ）模型外推、马尔科夫链等方法，研究了稻飞虱、稻纵卷叶螟、稻纹枯病、粘虫的长期运动规律。通过对这４种病虫１９９３～１９９５年的超长期预测研究，得出不同病虫适用的超长期预测方法有所不同，如稻纹枯病适宜用方差周期外推和特尔菲法，稻飞虱适宜用季节模型、时间序列和特尔菲法。而用上述适宜方法的预测值算术平均后的综合预报值预测的准确率更优于单一方法的预测。The paper analysed the statistical data of main diseases and insect pests on rice and wheat in Sichuan Province ,presented the grading level of occurrence and improved the Delphi,ARHMA model and Mlkefu et al. The paper also studied the long term occurrence of planthopper,rice leaf roller,rice blast and Mythimna separata. By studing the long term prediction of above four diseases and pests from 1993 to 1995, we found that different diseases and pests should use different ways,for example,the long term prediction of rice blast should use Delphi way and variance cycle extrapolating way,planthopper should use season model,time series and Delphi.The coincident rate of combining above methods are higher than any single way

关 键 词：统计资料 农作物病虫 超长期预测研究 

分 类 号：S431.2[农业科学—农业昆虫与害虫防治]