基于机器学习的保温被应用性能分析  

Performance analysis of insulation blanket application based on machine learning

作　　者：朱寅宾 骆乾亮 雷喜红[3] 牛曼丽 王平智[1,2] 程杰宇 赵淑梅[1,2] ZHU Yin-bin;LUO Qian-liang;LEI Xi-hong;NIU Man-li;WANG Ping-zhi;CHENG Jie-yu;ZHAO Shu-mei(College of Water Resources and Civil Engineering,China Agricultural University,Beijing 100083,China;Key Laboratory of Agricultural Engineering in Structure and Environment,Ministry of Agriculture and Rural Affairs,Beijing 100083,China;Beijing Agricultural Technology Extension Station,Beijing 100029,China)

机构地区：[1]中国农业大学水利与土木工程学院,北京100083 [2]农业农村部设施农业工程重点实验室,北京100083 [3]北京市农业技术推广站,北京100029

出　　处：《湖北农业科学》2025年第1期162-167,219,共7页Hubei Agricultural Sciences

基　　金：现代农业产业技术体系北京市设施蔬菜创新团队项目(BAIC01-2024)。

摘　　要：为满足装配式日光温室夜间保温需要以及研发新型温室保温材料,探索了机器学习在温室环境评价方面的应用,比较分析了骆驼绒和橡塑板为保温芯材的两种新型保温被保温性能。结果表明,高斯回归过程和神经网络算法在温室温度预测方面具有良好的应用潜力。相较于骆驼绒保温被,橡塑板保温被能使温室夜间薄膜内表面平均温度提高0.8℃,最低夜间薄膜内表面温度平均提高0.6℃。对于橡塑板芯材,应当加强防风措施管理以保证实际保温效果。To satisfy the nighttime insulation needs of prefabricated greenhouses and to develop novel insulation materials,the use of machine learning for evaluating greenhouse environments was investigated and the insulation efficacy of two new types of blankets was compared,one with camel hair and the other with rubber-plastic board as the core material.The findings indicated that both the Gauss-ian process regression and neural network algorithm held promise for predicting greenhouse temperatures.Compared to the camel hair blanket,the rubber-plastic insulation blanket increased the average night-time inner film surface temperature by 0.8℃and the aver-age minimum night-time temperature by 0.6℃.For the rubber-plastic board material,it was necessary to implement measures to miti-gate wind resistance in greenhouses to guarantee the insulation’s effectiveness.

关 键 词：保温被 薄膜内表面温度 机器学习 高斯过程回归 神经网络算法 

分 类 号：S626[农业科学—园艺学]