基于多元线性回归模型的日光温室温度预测方法对比  

Comparison of temperature prediction methods in greenhouse based on multilinear regression model

作　　者：冯大光[1] 冯思哲 李兆星 毛丽珍 FENG Daguang;FENG Sizhe;LI Zhaoxing;MAO Lizhen(Sciences of Institute,Shenyang Agricultural University,Shenyang 110866,China;School of Economics and Management,Southwest Jiaotong University,Chengdu 610031,China;College of Food Science,Shenyang Agricultural University,Shenyang 110866,China;Foreign Language Teaching Department,Shenyang Agricultural University,Shenyang 110866,China)

机构地区：[1]沈阳农业大学理学院,沈阳110866 [2]西南交通大学经济管理学院,成都610031 [3]沈阳农业大学食品学院,沈阳110866 [4]沈阳农业大学外语部,沈阳110866

出　　处：《沈阳师范大学学报(自然科学版)》2025年第1期75-81,共7页Journal of Shenyang Normal University:Natural Science Edition

基　　金：辽宁省教育厅普通高等教育本科教学改革研究项目(2020436)。

摘　　要：冬季北方日光温室温度是影响温室内作物生长的重要指标,温室温度的精准控制是作物正常生长的必要条件,而温度的准确预测是温度控制的前提。针对日光温室的温度预测问题,提出采用聚类分析和时间序列分析与多元线性回归模型相结合进行温度预测的方法,并以模型的误差平方和作为评价指标,对模型进行评价。结果表明,聚类分析和时间序列分析对多元线性回归模型的改进都是非常显著的,并且时间序列分析和多元线性回归相结合的模型更适合温室温度的预测。The greenhouse temperature in north China is an important indicator to the growth of vegetables in greenhouse,furthermore,the greenhouse temperature is one of the most important factors among all the controllable and uncontrollable factors,and the accurate prediction of temperature is the premise of temperature control.In this paper,a method of temperature prediction is proposed by combining K-means cluster and time series with multiple linear regressions,the sum of error square of the model is used as an evaluation index to evaluate the model.The results show that both K-means clusters method and time series method can improve the accuracy of prediction of multilinear regression with the objective function of total sum square of errors.And the models times series combined with multilinear regression model is the most efficient one,which can be easily carried out and applied to the control of greenhouse temperature.

关 键 词：温室 多元线性回归 K-均值聚类 时间序列分析 误差平方和 

分 类 号：S626.5[农业科学—园艺学] S646