我国进境植物疫情截获量的时序特征及预测  

Time series characteristics and prediction of plant pest intercepted in entry plant quarantine of China

作　　者：李瑞法[1] 柳之光 楚伟[1] 张学勤[1] 靳帅[1] 陈莹[1] 范光辉[1] 

机构地区：[1]河北出入境检验检疫局京唐港办事处,河北唐山063611

出　　处：《植物检疫》2016年第4期1-5,共5页Plant Quarantine

摘　　要：本文基于2005年1月-2015年10月月度植物疫情截获量数据和相应外贸进口额数据,利用回归分析和时间序列建模方法分析其时序特征并进行预测。结果表明,进口额每增长1亿美元,疫情截获量约增加47种次,进口额增长率每增长1个百分点,疫情截获量约增长2个百分点。疫情截获量时序存在显著的季节特征和指数型增长趋势,SARIMA模型和残差自回归模型可以较好地拟合疫情截获量时间序列并进行短期预测。最后根据研究结论提出了相关工作建议。Base on the data of the plant pest intercepted and corresponding volume of imports from Jan.,2005 to Oct.,2015,we get the time series characteristics,the prediction of the plant pest intercepted and the correlation with volume of imports with regression analysis and time series modeling method. It is concluded that the plant pest intercepted amount would increase 47 with a hundred million US dollars imports increase,and with per unit increase of the volume of imports growth rate,the growth rate of the plant pest intercepted increase two units. It was found that distinct seasonal characteristics and exponential population growth of the time series of the plant pest intercepted,SARIMA model and the Auto-Regressive model can fit and predict it correctly. Furthermore,working suggestions were proposed base on the above conclusion.

关 键 词：植物疫情 截获量 时间序列 季节自回归积分滑动平均模型 残差自回归 预测 

分 类 号：S41[农业科学—植物保护]