摘要
玉米是山东省主要的粮食作物之一,探究气象条件对山东省玉米产量的影响具有重要意义。本文采用山东省玉米生产相关数据,利用5a滑动平均法、双指数平滑法、回归分析法、HP滤波法对山东省玉米趋势产量进行拟合,并分离出相应的气象产量。利用SPSS相关分析和方差分析,对四种方法拟合趋势产量的显著程度和分离气象产量的准确程度进行比较。结果表明,四种方法均能较好地拟合趋势产量,但其分离气象产量的准确程度存在差异,准确性大小顺序为HP滤波分析>回归分析>5a滑动平均法>双指数平滑法。
Corn is one of the major food crops in Shandong Province. It has great significance to explore the influences of meteorological conditions on the yield of corn in Shandong Province. In this paper, the related data of corn production in Shandong Province were used to fit the trend yield by 5a sliding average method, double exponential smoothing method, regression analysis and HP filtering method respectively, and then the meteorological yield was separated out. By using SPSS correlation and variance analysis, the significant degree of the four methods for fitting the trend yield and the accuracy of separating meteorological yield were com- pared. The results showed that all the four methods could better fit the trend yield, but the accuracy of meteor- ological yield separation was different. The accuracy of the four methods was ordered as HP filtering analysis 〉 regression analysis 〉 5a sliding average method 〉 double exponential smoothing method.
出处
《山东农业科学》
2015年第8期95-99,共5页
Shandong Agricultural Sciences
基金
国家自然科学基金项目(71173139)
山东省社会科学规划研究项目(10CJGZ52)
关键词
山东省
玉米
趋势产量
气象产量
移动平均
回归分析
HP滤波
Shandong Province
Corn
Trend yield
Meteorological yield
Moving average
Regression analysis
HP filtering
作者简介
牛浩（1990-），男，山东莱芜人，博士研究生，研究方向为农业大数据资源利用。E-mail：niuha090@126．com
通讯作者：陈盛伟（1971-），男，山东青州人，管理学博士，教授，博士生导师，研究方向为农业保险。E—mail：chensw@sdau．edu．cn
