생물환경조절학회지 (Journal of Bio-Environment Control)

  • 제32권4호
  • /
  • Pages.384-395
  • /
  • 2023
  • /
  • 1229-4675(pISSN)
  • /
  • 2765-3641(eISSN)
한국생물환경조절학회 (The Korean Society for Bio-Environment Control)
권호 표지
DOI QR코드

DOI QR Code

라이시미터 데이터로 학습한 수학적 및 심층 신경망 모델을 통한 온실 토마토 증산량 추정

Estimation of Greenhouse Tomato Transpiration through Mathematical and Deep Neural Network Models Learned from Lysimeter Data

  • 메안 P 안데스 (필리핀 식물산업국 국립 작물연구개발 및 생산지원센터(BPI-LBNCRDPSC) ) ;
  • 노미영 (국립원예특작과학원 시설원예연구소 ) ;
  • 임미영 (국립원예특작과학원 시설원예연구소) ;
  • 최경이 (국립원예특작과학원 시설원예연구소) ;
  • 정정수 (경상북도 농업기술원 ) ;
  • 김동필 (국립원예특작과학원 시설원예연구소)
  • Meanne P. Andes (Bureau of Plant Industry-Los Baños National Crop Research Development and Production Support Center ) ;
  • Mi-young Roh (Protected Horticulture Research Institute, NIHHS, RDA) ;
  • Mi Young Lim (Protected Horticulture Research Institute, NIHHS, RDA) ;
  • Gyeong-Lee Choi (Protected Horticulture Research Institute, NIHHS, RDA) ;
  • Jung Su Jung (Gyeongsangbuk-do Provincial Agricultural Research and Extension Services) ;
  • Dongpil Kim (Protected Horticulture Research Institute, NIHHS, RDA)
  • 투고 : 2023.10.06
  • 심사 : 2023.10.24
  • 발행 : 2023.10.31
PDF 다운로드
⟨ 이전 논문 다음 논문 ⟩

초록

증산은 적정 관수 관리에 중요한 역할을 하므로 수분 스트레스에 취약한 토마토와 같은 작물의 관개 수요에 대한 지식이 필요하다. 관수량을 결정하는 한 가지 방법은 증산량을 측정하는 것인데, 이는 환경이나 생육 수준의 영향을 받는다. 본 연구는 분단위 데이터를 통해 수학적 모델과 딥러닝 모델을 활용하여 토마토의 증발량을 추정하고 적합한 모델을 찾는 것을 목표로 한다. 라이시미터 데이터는 1분 간격으로 배지무게 변화를 측정함으로써 증산량을 직접 측정했다. 피어슨 상관관계는 관찰된 환경 변수가 작물 증산과 유의미한 상관관계가 있음을 보여주었다. 온실온도와 태양복사는 증산량과 양의 상관관계를 보인 반면, 상대습도는 음의 상관관계를 보였다. 다중 선형 회귀(MLR), 다항 회귀 모델, 인공 신경망(ANN), Long short-term memory(LSTM), Gated Recurrent Unit(GRU) 모델을 구축하고 정확도를 비교했다. 모든 모델은 테스트 데이터 세트에서 0.770-0.948 범위의 R2 값과 0.495mm/min-1.038mm/min의 RMSE로 증산을 잠재적으로 추정하였다. 딥러닝 모델은 수학적 모델보다 성능이 뛰어났다. GRU는 0.948의 R2 및 0.495mm/min의 RMSE로 테스트 데이터에서 최고의 성능을 보여주었다. LSTM과 ANN은 R2 값이 각각 0.946과 0.944, RMSE가 각각 0.504m/min과 0.511로 그 뒤를 이었다. GRU 모델은 단기 예측에서 우수한 성능을 보였고 LSTM은 장기 예측에서 우수한 성능을 보였지만 대규모 데이터 셋을 사용한 추가 검증이 필요하다. FAO56 Penman-Monteith(PM) 방정식과 비교하여 PM은 MLR 및 다항식 모델 2차 및 3차보다 RMSE가 0.598mm/min으로 낮지만 분단위 증산의 변동성을 포착하는 데 있어 모든 모델 중에서 가장 성능이 낮다. 따라서 본 연구 결과는 온실 내 토마토 증산을 단기적으로 추정하기 위해 GRU 및 LSTM 모델을 권장한다.

Since transpiration plays a key role in optimal irrigation management, knowledge of the irrigation demand of crops like tomatoes, which are highly susceptible to water stress, is necessary. One way to determine irrigation demand is to measure transpiration, which is affected by environmental factor or growth stage. This study aimed to estimate the transpiration amount of tomatoes and find a suitable model using mathematical and deep learning models using minute-by-minute data. Pearson correlation revealed that observed environmental variables significantly correlate with crop transpiration. Inside air temperature and outside radiation positively correlated with transpiration, while humidity showed a negative correlation. Multiple Linear Regression (MLR), Polynomial Regression model, Artificial Neural Network (ANN), Long short-term Memory (LSTM), and Gated Recurrent Unit (GRU) models were built and compared their accuracies. All models showed potential in estimating transpiration with R2 values ranging from 0.770 to 0.948 and RMSE of 0.495 mm/min to 1.038 mm/min in the test dataset. Deep learning models outperformed the mathematical models; the GRU demonstrated the best performance in the test data with 0.948 R2 and 0.495 mm/min RMSE. The LSTM and ANN closely followed with R2 values of 0.946 and 0.944, respectively, and RMSE of 0.504 m/min and 0.511, respectively. The GRU model exhibited superior performance in short-term forecasts while LSTM for long-term but requires verification using a large dataset. Compared to the FAO56 Penman-Monteith (PM) equation, PM has a lower RMSE of 0.598 mm/min than MLR and Polynomial models degrees 2 and 3 but performed least among all models in capturing variability in transpiration. Therefore, this study recommended GRU and LSTM models for short-term estimation of tomato transpiration in greenhouses.

키워드

과제정보

This work was carried out with the support of "Cooperative Research Program for Agriculture Science and Technology Development (Project No. PJ01604801)" Rural Development Administration, Republic of Korea. We would also like to acknowledge the KOPIA Philippine Center for their unwavering support in enhancing the knowledge of trainees from partner country through provision of opportunities to attend long term training course and conduct this study.

참고문헌

  1. Allen R.G., L.S. Pereira, D. Raes, and M. Smith 1998, Crop evapotranspiration-guidelines for computing crop water requirements. FAO, Irrigation and Drainage Paper 56. Rome, Italy, p 300.
  2. Allen R.G., W.O. Pruitt, J.L. Wright, T.A. Howell, F. Ventura, R. Snyder, D. Itenfisu, P. Steduto, J. Berengena, J.B. Yrisarry, M. Smith, L.S. Pereira, D. Raes, A. Perrier, I. Alves, I. Walter, and R. Elliot 2006, A recommendation on standardized surface resistance for hourly calculation of reference Eto by the FAO56 Penman-Monteith method. Agric Water Manag 81:1-22. doi:10.1016/j.agwat.2005.03.007
  3. Bejo S.K., S. Mustaffha, and W.I.W. Ismail 2014, Application of Artificial Neural Network in predicting crop yield: A review. J Food Eng 4:1-9.
  4. Bera D., N.D. Chatterjee, and S. Bera 2021, Comparative performance of linear regression, polynomial regression and generalized additive model for canopy cover estimation in the dry deciduous forest of West Bengal. Remote Sens Appl Soc Environ 22:100502. doi:10.1016/j.rsase.2021.100502
  5. Chauhan N.S. 2020, Optimization algorithms in Neural Networks. KDNuggets. Available via https://www.kdnuggets.com/2020/12/optimization-algorithms-neural-networks.html Accessed August 28 2023.
  6. Chen Z., Z. Zhu, H. Jiang, and S. Sun 2020, Estimating daily reference evapotranspiration based on limited meteorological data using deep learning and classical machine learning methods. J Hydrol 591:125286. doi:10.1016/j.jhydrol.2020.125286
  7. Chia M.Y., Y.F. Huang, C.H. Koo, J.L. Ng, A.N. Ahmed, and A. El-Shafie 2022, Long-term forecasting of monthly mean reference evapotranspiration using Deep Neural Network: A comparison of training strategies and approaches. Appl Soft Comput 126:109221. doi:10.1016/j.asoc.2022.109221
  8. Dahikar S., and S. Rode 2014, Agricultural crop yield prediction using artificial neural network approach. Int J Innov Res Electric Electron Instrum Control Eng 2:683-686.
  9. De Wit C.T. 1958. Transpiration and Crop Yields. Institute of biological and chemical research of field crops and herbage. Wageningen 60:29-31. doi:10.1002/csc2.20094
  10. Elbeltagi A., N. Kumari, J.K. Dharpure, A. Mokhtar, K. Alsafadi, M. Kumar, B. Mehdinejadiani, H.R. Etedali, Y. Brouziyne, A.R.M.T. Islam, and A. Kuriqi 2021, Prediction of combined terrestrial evapotranspiration index (CTEI) over large river basin based on machine learning approaches. Water 13:547. doi:10.3390/w13040547
  11. Fan J., J. Zheng, L. Wu, and F. Zhang 2021, Estimation of daily maize transpiration using support vector machines, extreme gradient boosting, artificial and deep neural network models. Agric Water Manag 245:106547. doi:10.1016/j.agwat.2020.106547
  12. Fernandez J.E., and M.V. Cuevas 2010, Irrigation scheduling from stem diameter variations: a review. Agric For Meteorol 150:135-151. doi:10.1016/j.agrformet.2009.11.006
  13. Fernandez M.D., S. Bonachela, F. Orgaz, R.B. Thomson, J.C. Lopez, M.R. Granados, M. Gallardo, and E. Fereres 2011, Erratum to: Measurement and estimation of plastic greenhouse reference evapotranspiration in Mediterranean climate. Irrig Sci 29:91-92. doi:10.1007/s00271-010-0210-z
  14. Ferreira L.B., and F.F. da Cunha 2020, New Approach to estimate daily reference evapotranspiration based on hourly temperature and relative humidity using machine learning and deep learning. Agric Water Manag 234:106113. doi:10.1016/j.agwat.2020.106113
  15. Gallardo M., M.D. Fernandez, C. Gimenez, F.M. Padilla, and R.B. Thompson 2016, Revised VegSyst model to calculate dry matter production, critical N uptake and ETc of several vegetable species grown in Mediterranean greenhouses. Agric Syst 146:30-43. doi:10.1016/j.agsy.2016.03.014
  16. Gallardo M., R.B. Thompson, and M.D. Fernandez 2013, Water requirements and irrigation management in Mediterranean greenhouses: the case of the southeast coast of Spain. In W Baudoin, R Nono-Womdim, N Lutaladio, A Hodder, N Castilla, C Leonardi, S de Pascale, and M. Qaryouti, eds, Good Agriculture Practices for Greenhouse Vegetable Crops: Principles of Mediterranean Climate Areas. FAO, Rome, Italy, pp 109-136.
  17. Geelen P.A.M., J.O. Voogt, and P.A. van Weel 2020, Plant empowerment: The basic principles. Ed 2. Letsgrow.com.
  18. Hazlett D. 2022, Importance of transpiration rates. J Glob Sci Res 10:7-8. doi:10.15651/GJARR.22.9.4
  19. Incrocci L., R.B. Thompson, M.D.F. Fernandez, S. de Pascale, A. Pardossi, C. Stanghellini, Y. Rouphael, and M. Gallardo 2020, Irrigation management of European greenhouse vegetable crops. Agric Water Manag 242:106393. doi:10.1016/j.agwat.2020.106393
  20. Jo W.J., and J.H. Shin 2021, Development of a transpiration model for precise tomato (Solanum lycopersicum L.) irrigation control under various environmental conditions. Plant Physiol Biochem 162:388-394. doi:10.1016/j.plaphy.2021.03.005
  21. Jolliet O., and B.J. Bailey 1992, The effect if climate on tomato transpiration in greenhouses: measurements and models comparison. Agric For Meteorol 58:43-62. doi:10.1016/0168-1923(92)90110-P
  22. Katsoulas N., and C. Stanghellini 2019, Modelling crop transpiration in greenhouses: different models for different applications. Agronomy 9:392. doi:10.3390/agronomy9070392
  23. Li Y., J. Ye, D. Xu, G. Zhou, and H. Feng 2022, Prediction of sap flow with historical environmental factors based on deep learning technology Comput Electron Agric 202:107400. doi:10.1016/j.compag.2022.107400
  24. Li Y., L. Gou, J. Wang, Y. Wang, D. Xu, and J. Wen 2023, An improved sap flow prediction model based on CNN-GRUBiLSTM and factor analysis of historical environmental variables. Forests 14:1310. doi:10.3390/f14071310
  25. Magan J.J., M. Gallardo, R.B. Thompson, and P. Lorenzo 2008, Effects of salinity on fruit yield and quality of tomato grown in soil-less culture in greenhouses in Mediterranean climatic conditions. Agric Water Manag 95:1041-1055. doi:10.1016/j.agwat.2008.03.011
  26. Mehdizadeh S., B. Mohammadi, Q.B. Pham, and Z. Duan 2021, Development of boosted machine learning models for estimating daily reference evapotranspiration and comparison with empirical approaches. Water 13:3489. doi:10.3390/w13243489
  27. Mohammadi B., and S. Mehdizadeh 2020, Modeling daily reference evapotranspiration via a novel approach based on support vector regression coupled with whale optimization algorithm. Agric Water Manag 237:106145. doi:10.1016/j.agwat.2020.106145
  28. Mohammadi B., R. Moazenzadeh, K. Christian, Z., and Duan 2021, Improving streamflow simulation by combining hydrological process-driven and artificial intelligence-based models. Environ Sci Pollut Res 28:65752-65768. doi:10.1007/s11356-021-15563-1
  29. Nam D.S., T. Moon, J.W. Lee, and J.E. Son 2019, Estimating transpiration rates of hydroponically-grown paprika via an artificial neural network using aerial and root-zone environments and growth factors in greenhouses. Hortic Environ Biotechnol 60:913-923. doi:10.1007/s13580-019-00183-z
  30. Nwogu E.C., I.S. Iwueze, and V.U. Nlebedim 2016, Some tests for seasonality in time series data. J Mod Appl Stat Methods 15:382-399. doi:10.22237/jmasm/1478002920
  31. Ostertagova E. 2012, Modelling using polynomial regression. Procedia Eng 48:500-506. doi:10.1016/j.proeng.2012.09.545
  32. Park Y.S., and S. Lek 2016, Chapter 7-artificial neural networks: Multilayer perceptron for ecological modeling. Dev Environ Model 28:123-140. doi:10.1016/B978-0-444-63623-2.00007-4
  33. PASSeL (Plant and Soil Science eLibrary) 2023, Transpiration-factors affecting rates of transpiration. https://passel2.unl.edu/view/lesson/c242ac4fbaaf/6. Accessed August 28 2023
  34. Priestly C.H.B., and R.J. Taylor 1972, On the assessment of surface heat flux and evaporation using large-scale parameters. Mon Weather Rev 100:81-92. doi:10.1175/1520-0493(1972)100<0081:OTAOSH>2.3.CO;2
  35. Saedi R. 2022, Evaluation of multivariate regression models in estimation of evaporation and transpiration components of maize, under salinity stress conditions. Iran J Soil Water Res 53:71-84. doi:10.22059/ijswr.2022.335453.669157
  36. Sanchez J.A., F. Rodriguez, J.L. Guzman, and M.R. Arahal 2012, Virtual sensors for designing irrigation controllers in greenhouses. Sensors 12:15244-15266. doi:10.3390/s121115244
  37. Shao M., H.Liu, and L.Yan 2022, Estimating tomato transpiration cultivated in a sunken solar greenhouse with the Penman-Monteith, Shuttleworth-Wallace and Priestly-Taylor models in the North China plain. Agronomy 12:2382. doi:10.3390/agronomy12102382
  38. Shin J.H., and J.E. Son 2015, Development of a real-time irrigation control system considering transpiration, substrate electrical conductivity, and drainage rate of nutrient solutions in soilless culture of paprika (Capsicum annuum L.). Eur J Hortic Sci 80:271-279. doi:10.17660/eJHS.2015/80.6.2
  39. Shin J.H., J.S. Park, and J.E. Son 2014, Estimating actual transpiration rate with compensated levels of accumulated radiation for the efficient irrigation of soilless cultures of paprika plants. Agric Water Manag 135:9-18. doi:10.1016/j.agwat.2013.12.009
  40. Shuttleworth W.J., and J.S. Wallace 1985, Evaporation from sparse crops-An energy combination theory. Q J R Meteorol 111:839-855. doi:10.1002/qj.49711146910
  41. Tian H., P. Wang, K. Tansey, J. Zhang, S. Zhang, and H. Li 2020, An LSTM neural network for improving wheat yield estimates by integrating remote sensing data and meteorological data in the Guanzhong Plain, PR China. Agric For Meteorol 310:108629. doi:10.1016/j.agrformet.2021.108629
  42. Tranmer M., J. Murphy, M. Elliot, and M. Pampaka 2020, Multiple linear regression (2nd Edition). Cathie March institute working paper. https://hummedia.manchester.ac.uk/institutes/cmist/archive-publications/working-papers/2020/multiple-linear-regression.pdf Accessed August 28 2023
  43. Tu J., X. Wei, B. Huang, H. Fan, M. Jian, and W. Li 2019, Improvement of sap flow estimation by including phonological index and time-lag effect in back-propagation neural network models. Agric For Meteorol 276:107608. doi:10.1016/j.agrformet.2019.06.007
  44. Tunali U., I.H. Tuzel, Y. Tuzel, and Y. Senol 2023, Estimation of actual crop evapotranspiration using artificial neural networks in tomato grown in closed soilless culture system. Agric Water Manag 284:108331. doi:10.1016/j.agwat.2023.108331
  45. Vanja S., M. Eibl, and C. Hochenauer 2021, Artificial intelligence for time-efficient prediction and optimization of solid oxide fuel performances. Energy Convers Manag 230:113764. doi:10.1016/j.enconman.2020.113764.
  46. Yildirim E., and M. Ekinci 2022, Vegetable crops: health benefits and cultivation. IntechOpen:95704. doi:10.5772/intechopen.95704
  47. Ying X. 2019, An overview of overfitting and its solutions. J Phys Conf Ser 1168:022022. doi:10.1088/1742-6596/1168/2/022022
  48. Yong S.L.S., J.L. Ng, Y.F. Huang, and C.K. Ang 2023, Estimation of reference crop evapotranspiration with three different machine learning models and limited meteorological variables. Agronomy 13:1048. doi:10.3390/agronomy13041048
  49. Zarzycki K., and M. Lawrynczuk 2021, LSTM and GRU neural networks as models of dynamical processes used in predictive control: A comparison of models developed for two chemical reactors. Sensors 21:5625. doi:10.3390/s21165625
  50. Zhang J., Y. Zhu, X. Zhang, M. Ye, and J. Yang 2018, Developing a long short-term memory (LSTM) based model for predicting water table depth in agricultural areas. J Hydrol 561:918-929. doi:10.1016/j.jhydrol.2018.04.065
  51. Zhu Y., Z. Cheng, K. Feng, C. Zhang, C. Cao, J. Huang, H. Ye, and Y. Gao 2022. Influencing factors for transpiration rate: A numerical simulation of an individual leaf system. Therm Sci Eng Prog 27:101110. doi:10.1016/j.tsep.2021.101110