English

中国农机化学报

中国农机化学报 ›› 2025, Vol. 46 ›› Issue (6): 176-182.DOI: 10.13733/j.jcam.issn.2095-5553.2025.06.026

• 设施农业与植保机械工程 • 上一篇    下一篇

基于空间插值的温室三维温度场构建方法比较研究

夏皓1,2,刘杨1,2,贡宇2,任妮2,金晶2   

  1. (1. 江苏大学农业工程学院,江苏镇江,212013; 2. 江苏省农业科学院农业信息研究所/农业农村部长三角智慧农业技术重点实验室,南京市,210014)

  • 出版日期:2025-06-15 发布日期:2025-05-22
  • 基金资助:
    江苏省333人才项目(2022);江苏省农业科技自主创新资金项目(CX(22)5006)

Comparison study on the construction methods of three‑dimensional temperature field in a greenhouse based on spatial interpolation

Xia Hao1, 2, Liu Yang1, 2, Gong Yu2, Ren Ni2, Jin Jing2   

  1. (1. School of Agricultural Engineering, Jiangsu University, Zhenjiang, 212013, China; 2. Institute of Agricultural Information, Jiangsu Academy of Agricultural Sciences/Key Laboratory of Intelligent Agricultural Technology (Yangtze River Delta), Ministry of Agriculture and Rural Affairs, Nanjing, 210014, China)

  • Online:2025-06-15 Published:2025-05-22

摘要:

为确定适用于温室三维温度场构建的空间插值方法及插值分辨率,基于Venlo型玻璃温室中布设的27个温度传感器观测数据,使用反距离权重插值(IDW)、径向基函数插值(RBF)及普通克里金插值(OK)3种空间插值方法构建温室三维温度场,分析插值精度,比较8种不同空间插值分辨率对插值效率的影响。结果表明,IDW、RBF、OK的温度预测值与实测值均呈显著正相关关系(P<0.05,R2分别为0.928、0.911、0.957),平均绝对误差和均方根误差分别为0.766 ℃、0.765 ℃、0.599 ℃和1.266 ℃、1.449 ℃、0.981 ℃,其中OK构建的三维温度场效果最优。此外,与1 cm插值分辨率相比,20 cm分辨率下预测值的均方根误差为0.025 ℃,可在保证插值精度的条件下,减少99.96%的运算时间,可作为适宜的空间插值分辨率进行温度场构建。

关键词: 三维温度场, 传感器, 空间插值, 空间分辨率, 温室

Abstract:

 In order to determine the spatial interpolation method and interpolation resolution suitable for the construction of the 3D temperature field of the greenhouse, based on the observation data of 27 temperature sensors installed in the Venlo type glass greenhouse, three spatial interpolation methods, namely inverse distance weight interpolation (IDW), radial basis function interpolation (RBF) and ordinary Kriging interpolation (OK), were used to construct the 3D temperature field of the greenhouse. The interpolation accuracy was analyzed, and the effect of 8 different spatial interpolation resolutions on the interpolation efficiency was compared. The results showed that the predicted values of IDW, RBF and OK were significantly positively correlated with the measured values (P<0.05, R2 were 0.928, 0.911, 0.957, respectively). The mean absolute errors and root mean square error were 0.766 ℃, 0.765 ℃, 0.599 ℃ and 1.266 ℃, 1.449 ℃, 0.981 ℃, respectively. Among them, the three‑dimensional temperature field constructed by OK has the best effect. In addition, compared with the interpolation resolution of 1 cm, the root mean square error of the predicted value at the resolution of 20 cm was 0.025 ℃, which can reduce the operation time by 99.96% under the condition of ensuring the interpolation accuracy, and can be used as a suitable spatial interpolation resolution for temperature field construction.

Key words: three?dimensional temperature field, sensor, spatial interpolation, spatial resolution, greenhouse

中图分类号: