马尾松毛虫发生面积预测方法分析研究  

Analysis and Study on Prediction Method of Occurrence Area of Dendrolimus Punctatus

作　　者：邱利军[1] 张波[1] 杨怀义[2] 侯占东 QIU Li-jun;ZHANG Bo;YANG Huai-yi;HOU Zhan-dong(Hebei Institute of Architecture and Civil Engineering,Zhangjiakou,Hebei 075000;Shanxi Branch of China National Geological Exploration Center of Building Materials Industry,Taiyuan,Shanxi 030031;Jiangsu Geology Engineering Research Institute Co.,Ltd.,China Coal Group Changzhou,Jiangsu 213018)

机构地区：[1]河北建筑工程学院,河北张家口075000 [2]中国建筑材料工业地质勘查中心山西总队,山西太原030031 [3]江苏中煤地质工程研究院有限公司,江苏常州213018

出　　处：《河北建筑工程学院学报》2020年第4期69-71,75,共4页Journal of Hebei Institute of Architecture and Civil Engineering

基　　金：张家口市2016年度市级科技计划自筹经费项目(1621013B);张家口市2018年度市级科技计划自筹经费项目(1821019B)。

摘　　要：根据浙江省仙居县1983年至2001年马尾松毛虫历史资料,应用多元线性回归方法对发生面积进行预测,并对数据累计序列、新陈代谢序列和数据优化序列建模预测,并对其进行了比较分析。结果表明,数据优化序列建模的多元线性回归模型的预测结果最优,平均误差在10%,达到了较为满意的结果。因此,认为采用多元线性回归模型对马尾松毛虫发生面积进行预测能够达到较好效果,而且对森林保护有一定的实际应用价值和意义。Based on the historical data of dendrolimus punctatus in Xianju county of Zhejiang Province from 1983 to 2001,the multiple linear regression method is used to forecast the occurrence area of dendrolimus punctatus,and the data accumulation sequence,metabolism sequence and data optimization sequence are modeled to predict.Through comparison and analysis,the results show that the prediction result of multiple linear regression model based on data optimization sequence model is the best,and the average error is 10%which achieves satisfactory results.Therefore,multiple linear regression model can be used to predict dendrolimus punctatus,and it has a certain practical value and significance for forest protection.

关 键 词：多元线性回归 马尾松毛虫 发生面积 预测 

分 类 号：S763[农业科学—森林保护学]