基于门控循环单元的基质栽培黄瓜结果期蒸散量预测模型

朱鑫; 林琼; 何淽琦; 易志刚

基于门控循环单元的基质栽培黄瓜结果期蒸散量预测模型

朱鑫^{1, 2,},
林琼^2, ,,
何淽琦²,
易志刚^{1, 3}

1.
福建农林大学机电工程学院，福建　福州　350002
2.
福建省农业科学院资源环境与土壤肥料研究所，福建　福州　350013
3.
福建农林大学资源与环境学院，福建　福州　350002

基金项目: 福建省科技计划区域发展项目（2021N3008）；福建省科技计划公益类专项（2020R1025006）

详细信息

作者简介:
朱鑫（1997 —），男，硕士研究生，主要从事机电一体化研究，E-mail：18438606867@163.com

通讯作者:
林琼（1972 —），男，副研究员，主要从事植物营养与无土栽培研究，E-mail：linqiong@163.com

中图分类号: S161.4+2;TP312
计量
- 文章访问数: 6
- HTML全文浏览量: 4
- PDF下载量: 0
- 被引次数: 0
出版历程
- 收稿日期: 2023-12-21
- 录用日期: 2024-05-22
- 修回日期: 2024-01-15
- 网络出版日期: 2024-06-26

Prediction Model of Evapotranspiration during the Fruit Period of Cucumber Grown in Substrate Based on Gated Recurrent Unit

ZHU Xin^{1, 2
,},
LIN Qiong^{2
, ,},
HE Zhiqi²,
YI Zhigang^{1, 3}

1.
College of Mechanical and Electrical Engineering, Fujian Agriculture and Forestry University, Fujian Fuzhou　350002, China
2.
Institute of Resources, Environment and Soil Fertilizer, Fujian Academy of Agricultural Sciences, Fujian Fuzhou　350013, China
3.
College of Resources and Environment, Fujian Agriculture and Forestry University, Fujian Fuzhou　350002, China

摘要

摘要: 目的实时、准确地预测基质栽培黄瓜结果期蒸散量，指导基质栽培黄瓜灌溉。方法通过传感器实时获取黄瓜结果期的温室小气候环境数据，用称量法测量黄瓜蒸散量，以移栽时间、空气温度、空气相对湿度、光照强度及前5天的日均灌溉量为输入，利用BP神经网络（Back Propagation Neural Network, BPNN）、卷积神经网络（Convolutional Neural Networks, CNN）、长短期记忆网络（Long Short-Term Memory, LSTM）和门控循环单元（Gated recurrent unit, GRU）分别建立基质栽培黄瓜蒸散量预测模型，比较不同模型的预测效果，模型数据集的时间间隔设为20 min。结果相较于BPNN、CNN及LSTM模型，GRU模型的预测效果最好，其决定系数（R²）、均方根误差（RMSE）、平均绝对误差（MAE）分别为0.8577、2.3279 g和1.6744 g。当实测的黄瓜每日实时累积蒸散量超过50 g时，GRU模型预测的黄瓜每日实时累积蒸散量与实测每日实时累积蒸散量之间的相对误差波动最小，在0.11%～10.01%。结论基于GRU的基质栽培黄瓜结果期蒸散量预测模型预测效果最好，可为基质栽培黄瓜的灌溉系统提供参考。
- 蒸散量 /
- 基质栽培 /
- 门控循环单元 /
- 预测模型
Abstract: :[ Objective ] Real time and accurate prediction of evapotranspiration during the fruiting period of substrate cultivated cucumber, guiding substrate cultivated cucumber irrigation. Method Real time greenhouse microclimate environmental data during the cucumber fruiting period is obtained through sensors. Cucumber evapotranspiration is measured using the weighing method, with transplanting time, air temperature, relative humidity, light intensity, and daily average irrigation amount for the first five days as inputs, using Back Propagation neural network (BPNN), Convolutional Neural Networks (CNN), Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU) to establish substrate cultivation cucumber evapotranspiration prediction models, and the prediction effects of different models are compared. The model dataset The time interval is set to 20 minutes. Result Compared to BPNN, CNN, and LSTM models, the GRU model has the best prediction performance, with a coefficient of determination (R²), root mean square error (RMSE), and mean absolute error (MAE) of 0.8577, 2.3279 g, and 1.6744 g, respectively. When the measured daily real-time cumulative evapotranspiration of cucumbers exceeds 50 g, the relative error fluctuation between the predicted daily real-time cumulative evapotranspiration of cucumbers by the GRU model and the measured daily real-time cumulative evapotranspiration is the smallest, ranging from 0.11% to 10.01%. Conclusion The prediction model of evapotranspiration during the fruiting period of substrate cultivated cucumbers based on GRU has the best prediction effect, which can provide reference for irrigation system of substrate cultivated cucumbers.
- Evapotranspiration /
- Substrate cultivation /
- Gated recurrent unit /
- Prediction model

HTML全文

图 1 黄瓜叶面积、黄瓜日均蒸散量与黄瓜日均灌溉量的关系

Figure 1. The Relationship between Cucumber Leaf Area, Cucumber Daily Average Evapotranspiration, and Cucumber Daily Average Irrigation

下载: 全尺寸图片幻灯片

图 2 不同模型预测值与实测值的相关曲线

Figure 2. Correlation curves between predicted and measured values of different models

下载: 全尺寸图片幻灯片

图 3 黄瓜每日实时累积蒸散量曲线

Figure 3. Daily real-time cumulative evapotranspiration curve of cucumber

下载: 全尺寸图片幻灯片

表 1 GRU不同网络参数下的RMSE

Table 1. RMSE of GRU under different network parameters　　　　　　　　　　　　　　　　　　（单位：g）

时间步长 Time step	隐含层节点数 Number of hidden layer nodes
时间步长 Time step	5	10	15
12	2.4318	2.6083	2.6658
24	2.3279	2.3469	2.3374
36	2.4995	2.3547	2.4304

下载: 导出CSV

表 2 GRU不同网络参数下的MAE

Table 2. MAE of GRU under different network parameters 　　　　　　　　　　　　　　　　　　（单位：g）

时间步长 Time step	隐含层节点数 Number of hidden layer nodes
时间步长 Time step	5	10	15
12	1.6457	1.8546	1.9906
24	1.6744	1.6997	1.6951
36	1.8922	1.8152	1.8680

下载: 导出CSV

表 3 黄瓜每日实时累积蒸散量相对误差

Table 3. Relative error of daily real-time cumulative evapotranspiration of cucumbers　　　（单位：%）

模型 Model	日期 Date
模型 Model	2023-05-12	2023-05-13	2023-05-14
BP神经网络	23.50±20.03	36.69±28.72	17.51±16.35
CNN	8.73±8.73	7.74±7.59	14.27±14.01
LSTM	3.56±3.43	4.85±4.63	6.41±6.35
GRU	2.04±1.93	5.63±4.38	3.24±2.52

下载: 导出CSV

参考文献(28)

[1]	刘志雄, 陈磊夫, 戴照义, 等. 大棚黄瓜有机生态型无土栽培技术 [J]. 中国瓜菜, 2019, 32(11):97−98. doi: 10.3969/j.issn.1673-2871.2019.11.022 LIU Z X, CHEN L F, DAI Z Y, et al. Organic ecotype soilless culture technology of cucumber in greenhouse [J]. China Cucurbits and Vegetables, 2019, 32(11): 97−98. (in Chinese) doi: 10.3969/j.issn.1673-2871.2019.11.022
[2]	宋朝义. 黄瓜夏季育苗株型调控及栽培基质粒径研究[D]. 邯郸: 河北工程大学, 2020 SONG (C/Z)Y. Study on regulation of plant morphology of cucumber seedling in summer and the particle size of cultivation substrate[D]. Handan: Hebei University of Engineering, 2020. (in Chinese)
[3]	哈婷. 基质培黄瓜、番茄、茄子营养液供液制度研究[D]. 银川: 宁夏大学, 2017 HA T. Studies on supply system of nutrient solution on cucumber, tomato and eggplant under substrate cultivation[D]. Yinchuan: Ningxia University, 2017. (in Chinese)
[4]	李银坤, 郭文忠, 韩雪, 等. 基于称重式蒸渗仪实测值的温室茄子日蒸散量估算方法评价 [J]. 中国农业气象, 2020, 41(3):129−137. doi: 10.3969/j.issn.1000-6362.2020.03.001 LI Y K, GUO W Z, HAN X, et al. Evaluation of methods for estimating greenhouse eggplant daily evapotranspiration based on the values of weighing lysimeter measurements [J]. Chinese Journal of Agrometeorology, 2020, 41(3): 129−137. (in Chinese) doi: 10.3969/j.issn.1000-6362.2020.03.001
[5]	徐立鸿, 肖康俊, 蔚瑞华. 基于温室环境和作物生长的番茄基质栽培灌溉模型 [J]. 农业工程学报, 2020, 36(10):189−196. doi: 10.11975/j.issn.1002-6819.2020.10.023 XU L H, XIAO K J, WEI R H. Irrigation models for the tomatoes cultivated in organic substrate based on greenhouse environment and crop growth [J]. Transactions of the Chinese Society of Agricultural Engineering, 2020, 36(10): 189−196. (in Chinese) doi: 10.11975/j.issn.1002-6819.2020.10.023
[6]	赵爽, 闫浩芳, 张川, 等. 基于改进的双作物系数模型与Priestley-Taylor模型估算温室黄瓜蒸散量 [J]. 排灌机械工程学报, 2023, 41(8):849−857. ZHAO S, YAN H F, ZHANG C, et al. Estimation of cucumber evapotranspiration in greenhouse based on improved dual crop coefficient model and Priestley-Taylor model [J]. Journal of Drainage and Irrigation Machinery Engineering, 2023, 41(8): 849−857. (in Chinese)
[7]	姚勇哲, 李建明, 张荣, 等. 温室番茄蒸腾量与其影响因子的相关分析及模型模拟 [J]. 应用生态学报, 2012, 23(7):1869−1874. YAO Y Z, LI J M, ZHANG R, et al. Greenhouse tomato transpiration and its affecting factors: Correlation analysis and model simulation [J]. Chinese Journal of Applied Ecology, 2012, 23(7): 1869−1874. (in Chinese)
[8]	冀泽宇. 外源多巴胺对黄瓜霜霉病的缓解效应及其机理研究[D]. 杨凌: 西北农林科技大学, 2022 JI Z Y. Alleviation effect and mechanism of exogenous dopamine on cucumber downy mildew[D]. Yangling: Northwest A & F University, 2022. (in Chinese)
[9]	李晨, 李王成, 赵自阳, 等. 宁夏引黄灌区几种参考作物蒸散量计算方法适用性及修正研究 [J]. 中国农村水利水电, 2019, (11):54−59,65. doi: 10.3969/j.issn.1007-2284.2019.11.011 LI C, LI W C, ZHAO Z Y, et al. The applicability and correction of calculation methods for several reference crop evapotranspiration in Ningxia Yellow River irrigation districts [J]. China Rural Water and Hydropower, 2019(11): 54−59,65. (in Chinese) doi: 10.3969/j.issn.1007-2284.2019.11.011
[10]	郑勇东, 赵全明, 张馨, 等. 基于Lora无线技术的多路盆栽植物蒸散测量系统研发 [J]. 节水灌溉, 2020, (3):77−84. doi: 10.3969/j.issn.1007-4929.2020.03.015 ZHENG Y D, ZHAO Q M, ZHANG X, et al. Development of multi-channel potted-plant evapotranspiration measurement system based on lora wireless technology [J]. Water Saving Irrigation, 2020(3): 77−84. (in Chinese) doi: 10.3969/j.issn.1007-4929.2020.03.015
[11]	王怡宁, 朱月灵. 蒸渗仪国内外应用现状及研究趋势 [J]. 水文, 2018, 38(1):81−85. doi: 10.3969/j.issn.1000-0852.2018.01.014 WANG Y N, ZHU Y L. Application status and research trend of domestic and foreign lysimeter [J]. Journal of China Hydrology, 2018, 38(1): 81−85. (in Chinese) doi: 10.3969/j.issn.1000-0852.2018.01.014
[12]	KACIRA M, LING P P. Design and development of an automated and non-contact sensing system for continuous monitoring of plant health and growth[J]. Transactions of the ASAE American Society of Agricultural Engineers, 2001, 44(4): 989-996. [PubMed]
[13]	李亮斌, 姜晟, 王卫星, 等. 基于无线传感器网络的农村供水厂水质监测节点的设计 [J]. 湖南农业大学学报(自然科学版), 2016, 42(2):212−216. LI L B, JIANG S, WANG W X, et al. Design of wireless sensor network node for monitoring water quality of rural water supply plant [J]. Journal of Hunan Agricultural University (Natural Sciences), 2016, 42(2): 212−216. (in Chinese)
[14]	ZHANG M H, LI X, WANG L L. An adaptive outlier detection and processing approach towards time series sensor data[J]. IEEE Access, 2019, 7: 175192-175212. [LinkOut]
[15]	HUANG S, YAN H F, ZHANG C, et al. Modeling evapotranspiration for cucumber plants based on the Shuttleworth-Wallace model in a Venlo-type greenhouse[J]. Agricultural Water Management, 2020, 228: 105861. [LinkOut]
[16]	AHMADI F, MEHDIZADEH S, MOHAMMADI B, et al. Application of an artificial intelligence technique enhanced with intelligent water drops for monthly reference evapotranspiration estimation[J]. Agricultural Water Management, 2021, 244: 106622. [LinkOut]
[17]	MOHAMMADI B, MEHDIZADEH S. Modeling daily reference evapotranspiration via a novel approach based on support vector regression coupled with whale optimization algorithm[J]. Agricultural Water Management, 2020, 237: 106145. [LinkOut]
[18]	HOCHREITER S, SCHMIDHUBER J. Long short-term memory[J]. Neural Computation, 1997, 9(8): 1735-1780. [LinkOut]
[19]	李莉, 李伟, 耿磊, 等. 基于RF-GRU的温室番茄结果前期蒸腾量预测方法 [J]. 农业机械学报, 2022, 53(3):368−376. doi: 10.6041/j.issn.1000-1298.2022.03.039 LI L, LI W, GENG L, et al. Prediction method of greenhouse tomato transpiration in early fruiting stage based on RF-GRU [J]. Transactions of the Chinese Society for Agricultural Machinery, 2022, 53(3): 368−376. (in Chinese) doi: 10.6041/j.issn.1000-1298.2022.03.039
[20]	李莉, 李文军, 马德新, 等. 基于LSTM的温室番茄蒸腾量预测模型研究 [J]. 农业机械学报, 2021, 52(10):369−376. doi: 10.6041/j.issn.1000-1298.2021.10.038 LI L, LI W J, MA D X, et al. Prediction model of transpiration of greenhouse tomato based on LSTM [J]. Transactions of the Chinese Society for Agricultural Machinery, 2021, 52(10): 369−376. (in Chinese) doi: 10.6041/j.issn.1000-1298.2021.10.038
[21]	GONG X W, LIU H, SUN J S, et al. Comparison of Shuttleworth-Wallace model and dual crop coefficient method for estimating evapotranspiration of tomato cultivated in a solar greenhouse[J]. Agricultural Water Management, 2019, 217: 141-153. [LinkOut]
[22]	高建昌, 郭广君, 国艳梅, 等. 平台扫描仪结合ImageJ软件测定番茄叶面积 [J]. 中国蔬菜, 2011, (2):73−77. GAO J C, GUO G J, GUO Y M, et al. Measuring plant leaf area by scanner and ImageJ software [J]. China Vegetables, 2011(2): 73−77. (in Chinese)
[23]	姜含露, 周利明, 马明, 等. 基于多条件时间序列的免耕播种机作业数据清洗方法 [J]. 农业机械学报, 2022, 53(1):85−91. doi: 10.6041/j.issn.1000-1298.2022.01.009 JIANG H L, ZHOU L M, MA M, et al. Data cleaning method of No-tillage seeder monitoring data based on multi-conditional time series [J]. Transactions of the Chinese Society for Agricultural Machinery, 2022, 53(1): 85−91. (in Chinese) doi: 10.6041/j.issn.1000-1298.2022.01.009
[24]	裴孝伯, 李世诚, 张福墁, 等. 温室黄瓜叶面积计算及其与株高的相关性研究 [J]. 中国农学通报, 2005, 21(8):80−82. doi: 10.3969/j.issn.1000-6850.2005.08.022 PEI X B, LI S C, ZHANG F M, et al. Study on leaf area calculation and its correlation with plant height of cucumber in greenhouse [J]. Chinese Agricultural Science Bulletin, 2005, 21(8): 80−82. (in Chinese) doi: 10.3969/j.issn.1000-6850.2005.08.022
[25]	李艳大, 孙滨峰, 曹中盛, 等. 基于作物生长监测诊断仪的双季稻叶面积指数监测模型 [J]. 农业工程学报, 2020, 36(10):141−149. doi: 10.11975/j.issn.1002-6819.2020.10.017 LI Y D, SUN B F, CAO Z S, et al. Model for monitoring leaf area index of double cropping rice based on crop growth monitoring and diagnosis apparatus [J]. Transactions of the Chinese Society of Agricultural Engineering, 2020, 36(10): 141−149. (in Chinese) doi: 10.11975/j.issn.1002-6819.2020.10.017
[26]	龙泽昊, 秦其明, 张添源, 等. 基于长短期记忆网络的冬小麦连续时序叶面积指数预测 [J]. 光谱学与光谱分析, 2020, 40(3):898−904. LONG Z H, QIN Q M, ZHANG T Y, et al. Prediction of continuous time series leaf area index based on long short-term memory network: A case study of winter wheat [J]. Spectroscopy and Spectral Analysis, 2020, 40(3): 898−904. (in Chinese)
[27]	徐立鸿, 肖康俊, 蔚瑞华. 基于温室环境和作物生长的番茄基质栽培灌溉模型 [J]. 农业工程学报, 2020, 36(10):189−196. doi: 10.11975/j.issn.1002-6819.2020.10.023 XU L H, XIAO K J, WEI R H. Irrigation models for the tomatoes cultivated in organic substrate based on greenhouse environment and crop growth [J]. Transactions of the Chinese Society of Agricultural Engineering, 2020, 36(10): 189−196. (in Chinese) doi: 10.11975/j.issn.1002-6819.2020.10.023
[28]	李时雨. 日光温室椰糠袋培条件下黄瓜的灌溉模式研究[D]. 泰安: 山东农业大学, 2018 LI S Y. Study on the irrigation mode of cucumber under the condition of culture coconut bag in solar greenhouse[D]. Taian: Shandong Agricultural University, 2018. (in Chinese)