臺灣博碩士論文加值系統

在氣候變遷之影響下，極端天氣之頻率上升危及人類糧食及水資源之問題日益嚴重，衍伸出許多問題。由於溫室環境較不受外在因子影響且容易管理之優勢，溫室內之精準農業逐漸成為趨勢。為了有效提高水資源之利用且減少水資源之浪費，本研究在溫室內種植玉女小番茄並裝設監控儀器收集數據，希望藉由結合天氣預報對作物生長期間之土壤溫度及體積水分含量做預測。
首先，分別採用物理模式(Hydrus-1D)、機器學習(隨機森林)及動態拓樸學(ICON)的方法，模擬溫室內土壤溫度及體積含水量隨時間變化之情形。結果發現，這三種模式在模擬階段皆表現相當，均可以有效模擬溫室內之土壤溫度及體積水分含量。更進一步利用天氣預報資料，結合三種模式，預測土壤溫度及體積水分含量。
在物理模式下，所預測之土壤溫度趨勢均相似；且預測體積水分含量也較為準確，但卻無法預測人為澆灌之水分含量上升。在動態拓樸學模式下，預測土壤溫度的效果不彰。以機器學習之模式進行預測，不論是在土壤溫度或者體積水分含量之變化，皆可以得到最好之結果。
本研究應用Hydrus-1D、ICON以及隨機森林模擬並預測溫室內土壤溫度及體積水分含量之變化，能提供溫室管理之決策依據，達到智慧農業或無人農業之目標，同時提高水資源使用效率。

Due to the impact of climate change and the rising frequency of extreme weather have caused the problem of food security and water resources. Precision agriculture in greenhouses is gradually becoming a trend because it is less affected by external environmental factors and easier to manage. In order to effectively improve the utilization of water resources and reduce wasting water, this study planted cherry tomatoes in the greenhouse and installed monitoring instruments to collect data. The soil temperature and volumetric water content during tomato growth can be predicted effectively by combining the weather forecast.
The Hydrus-1D, random forest method and ICON were used to simulate the change of soil temperature and volumetric water content in the greenhouse. The results of the simulation show good performances among the three models. Further combined with weather forecast data, three models were applied to predict soil temperatures and volumetric moisture contents.
Both of predicted soil temperature and volumetric water content by Hydrus-1D followed the trend of real measured data, but Hydrus-1D can’t predict the watering situation. The predicted performance of soil temperature by ICON was low. Nevertheless, both predictions by random forest method were better than others.
This study simulated and predicted the soil temperatures and volumetric water contents by applying Hydrus-1D, random forest method and ICON to improve greenhouse management and increase water-use efficiency.

口試委員會審定書 #
誌謝 i
中文摘要 ii
ABSTRACT iii
目錄 iv
圖目錄 vii
Chapter 1 緒論 1
1.1 研究動機 1
1.2 研究目的 3
1.3 研究流程與方法 3
Chapter 2 材料與方法 7
2.1 溫室內試驗 7
2.1.1 溫室配置 7
2.1.2 玉女番茄 9
2.1.3 介質土 9
2.1.4 監測儀器 11
2.2 保水曲線實驗 14
2.3 水力傳導度實驗 16
Chapter 3 模式與相關理論 18
3.1 物理模式 18
3.1.1 牛頓冷卻定律(Newton’s cooling law) 18
3.1.2 土壤熱傳方程式(Soil heat conduction equation) 18
3.1.3 理查方程式(Richards’ equation) 19
3.2 機器學習模式 21
3.2.1 決策樹(Decision tree) 22
3.2.2 整體學習(Ensemble learning) 24
3.2.3 隨機森林(Random Forest) 25
3.3 動態拓樸學模式 26
Chapter 4 結果與討論 33
4.1 溫室監測資料整理與分析 33
4.1.1 第一期監測資料 33
4.1.2 第二期監測資料 36
4.1.3 時滯(Time lag)現象與交互關聯(Cross-correlation)分析 39
4.1.4 植物生長與蒸發散量分析 42
4.2 物理模式模擬 47
4.2.1 牛頓冷卻定律模擬 47
4.2.2 Hydrus-1D模擬 49
4.3 隨機森林模擬 54
4.3.1 土壤溫度模擬 54
4.3.2 體積水分含量模擬 56
4.4 ICON模擬 59
4.5 結合天氣預報與物理模型預測 61
4.5.1 天氣預報之準確度 62
4.5.2 天氣溫度與室內溫度之關係 63
4.5.3 室內溫度與上、下界土表溫度之關係 64
4.5.4 Hydrus-1D模型預測 65
4.6 結合天氣預報與機器學習模型預測 66
4.6.1 天氣預報之準確度 66
4.6.2 天氣溫度與室內溫度之關係 67
4.6.3 機器學習模型之預測 69
4.7 結合天氣預報與ICON模型預測 70
4.7.1 天氣預報與監測資料之模型 70
Chapter 5 結論與建議 73
Chapter 6 參考文獻 74

Ahmad, S., A. Kalra, H. Stephen (2010), Estimating soil moisture using remote sensing data: A machine learning approach, Advances in Water Resources, 33(1), 69-80, doi: 10.1016/j.advwatres.2009.10.008.
Al-Kayssi, A. W., A. A. Al-Karaghouli, A. M. Hasson, S. A. Beker (1990), Influence of Soil Moisture Content on Soil Temperature and Heat Storage under Greenhouse Conditions, J Agr Eng Res, 45(4), 241-252, doi: 10.1016/S0021-8634(05)80152-0.
Alzubi, J., A. Nayyar, A. Kumar (2018), Machine Learning from Theory to Algorithms: An Overview, Journal of Physics: Conference Series, 1142, doi: 10.1088/1742-6596/1142/1/012012.
Amprako, J. L. (2016), The United Nations World Water Development Report 2015: Water for a Sustainable World, Future Food, 4(2), 64-65.
Besson, U. (2012), The History of the Cooling Law: When the Search for Simplicity can be an Obstacle, Sci Educ-Netherlands, 21(8), 1085-1110.
Bot, G. P. (1991), Physical modeling of greenhouse climate, IFAC Proceedings Volumes, 24(11), 7-12.
Breiman, L. (1996), Bagging predictors, Machine learning, 24(2), 123-140.
Chung, S. O., R. Horton (1987), Soil Heat and Water-Flow with a Partial Surface Mulch, Water Resour Res, 23(12), 2175-2186.
Cunha, J. B. (2003), Greenhouse climate models: an overview, paper presented at Efita 2003 conference.
Deardorff, J. W. (1978), Efficient prediction of ground surface temperature and moisture, with inclusion of a layer of vegetation, Journal of Geophysical Research, 83(C4), doi: 10.1029/JC083iC04p01889.
Dong, J. Z., S. C. Steele-Dunne, T. E. Ochsner, C. E. Hatch, C. Sayde, J. Selker, S. Tyler, M. H. Cosh, N. van de Giesen (2016), Mapping high-resolution soil moisture and properties using distributed temperature sensing data and an adaptive particle batch smoother, Water Resour Res, 52(10), 7690-7710.
Duarte-Galvan, C., I. Torres-Pacheco, R. G. Guevara-Gonzalez, R. J. Romero-Troncoso, L. M. Contreras-Medina, M. A. Rios-Alcaraz, J. R. Mian-Amaraz (2012), Review. Advantages and disadvantages of control theories applied in greenhouse climate control systems, Span J Agric Res, 10(4), 926-938, doi: 10.5424/sjar/2012104-487-11.
Gebbers, R., V. I. Adamchuk (2010), Precision agriculture and food security, Science, 327(5967), 828-831, doi: 10.1126/science.1183899.
Giannakis, D., A. J. Majda (2014), Data-driven methods for dynamical systems: Quantifying predictability and extracting spatiotemporal patterns, Wiley & Sons.
Gonda, L., C. Eduardo Cugnasca (2006), A Proposal of Greenhouse Control Using Wireless Sensor Networks, paper presented at Computers in Agriculture and Natural Resources, 4th World Congress Conference, Proceedings of the 24-26 July 2006 (Orlando, Florida USA) Publication Date 24 July 2006, ASABE, St. Joseph, MI.
Hansen, P. C. (1987), The Truncated Svd as a Method for Regularization, Bit, 27(4), 534-553.
Hodge, A., G. Berta, C. Doussan, F. Merchan, M. Crespi (2009), Plant root growth, architecture and function, Plant Soil, 321(1-2), 153-187, doi: 10.1007/s11104-009-9929-9.
Hu, H. G., L. H. Xu, R. H. Wei, B. K. Zhu (2011), Multi-Objective Control Optimization for Greenhouse Environment Using Evolutionary Algorithms, Sensors-Basel, 11(6), 5792-5807, doi: 10.3390/s110605792.
Karandish, F., J. Šimůnek (2016), A comparison of numerical and machine-learning modeling of soil water content with limited input data, J Hydrol, 543, 892-909, doi: 10.1016/j.jhydrol.2016.11.007.
Kelley, J. L. (2017), General topology, Dover edition. ed., xiv, 298 pages pp., Dover Publications, Inc., Mineola, New York.
Kotsiantis, S. B. (2013), Decision trees: a recent overview, Artificial Intelligence Review, 39(4), 261-283.
Lynch, J. (1995), Root Architecture and Plant Productivity, Plant Physiol, 109(1), 7-13, doi: DOI 10.1104/pp.109.1.7.
Mohri, M., A. Rostamizadeh, A. Talwalkar (2012), Foundations of Machine Learning, 480 pp., The MIT Press.
Murphy, A. H. (1993), What is a good forecast? An essay on the nature of goodness in weather forecasting, Weather and forecasting, 8(2), 281-293.
Murphy, K. P. (2012), Machine learning : a probabilistic perspective, xxix, 1067 p. pp., MIT Press, Cambridge, Mass.
Nelson, G., M. Rosegrant, J. Koo, R. Robertson, T. Sulser, T. Zhu, C. Ringler, S. Msangi, A. Palazzo, M. Batka, M. Magalhaes, V.-S. Rowena, M. Ewing, D. Lee (2009), Climate change: Impact on agriculture and costs of adaptation, International Food Policy Research Institute (IFPRI).
Philip, J. R. (1957), Evaporation, and Moisture and Heat Fields in the Soil, Journal of Meteorology, 14(4), 354-366, doi: 10.1175/1520-0469(1957)014<0354:Eamahf>2.0.Co;2.
Richards, L. A. (1931), Capillary conduction of liquids through porous mediums, Physics-J Gen Appl P, 1(1), 318-333.
Shandilya, S. G., M. Timme (2011), Inferring network topology from complex dynamics, New J Phys, 13(1), 013004.
Simunek, J. (1998), The Hydrus-1D software package for simulating the one-dimensional movement of water, heat, and multiple solutes in variably-saturated media.
van Genuchten, M. T. (1980), A Closed-form Equation for Predicting the Hydraulic Conductivity of Unsaturated Soils1, Soil Science Society of America Journal, 44(5), doi: 10.2136/sssaj1980.03615995004400050002x.
Wang, S., E. D. Herzog, I. Z. Kiss, W. J. Schwartz, G. Bloch, M. Sebek, D. Granados-Fuentes, L. Wang, J. S. Li (2018), Inferring dynamic topology for decoding spatiotemporal structures in complex heterogeneous networks, P Natl Acad Sci USA, 115(37), 9300-9305, doi: 10.1073/pnas.1721286115.
Wang, W.-X., R. Yang, Y.-C. Lai, V. Kovanis, M. A. F. Harrison (2011), Time-series–based prediction of complex oscillator networks via compressive sensing, EPL (Europhysics Letters), 94(4), 48006.
WEI, R.-j., X.-p. WANG, G.-r. CHANG, Z.-l. LI, X.-n. WANG, Y. XIANG (2001), The Air Temperature Variation inside Greenhouse Related to the Outside Temperature in Continuous Cloudy-weather and its Regulation [J], Agricultural Meteorology, 3.
Zhang, N., M. Wang, N. Wang (2002), Precision agriculture—a worldwide overview, Computers and Electronics in Agriculture, 36(2-3), 113-132, doi: 10.1016/s0168-1699(02)00096-0.
Zhou, Z.-H. (2012), Ensemble methods : foundations and algorithms, xiv, 222 p. pp., CRC Press, Boca Raton, FL.
呂政道 (2009), 台灣農民年齡與農耕戶勞動生產力之關係, 1-91 pp, 臺灣大學.
蔡義誌 (2008), 不飽和土壤水力傳導度與介質孔隙分佈關係之研究, 1-163 pp, 中興大學.