생물환경조절학회지 (Journal of Bio-Environment Control)

  • 제29권3호
  • /
  • Pages.277-284
  • /
  • 2020
  • /
  • 1229-4675(pISSN)
  • /
  • 2765-3641(eISSN)
한국생물환경조절학회 (The Korean Society for Bio-Environment Control)
권호 표지
DOI QR코드

DOI QR Code

인공지능 기반 온실 환경인자의 시간영역 추정

A Research about Time Domain Estimation Method for Greenhouse Environmental Factors based on Artificial Intelligence

  • 이정규 (충북대학교 바이오시스템공학과 대학원) ;
  • 오종우 (티젯 테크놀로지 코리아) ;
  • 조용진 (전북대학교 생물산업기계공학과) ;
  • 이동훈 (충북대학교 바이오시스템공학과)
  • Lee, JungKyu (Department of Biosystems Engineering, Chungbuk National University) ;
  • Oh, JongWoo (Teejet Technologies) ;
  • Cho, YongJin (Department of Bio-Industrial Machinery Engineering, Jeonbuk National, University) ;
  • Lee, Donghoon (Department of Biosystems Engineering, Chungbuk National University)
  • 투고 : 2019.12.26
  • 심사 : 2020.06.23
  • 발행 : 2020.07.30
PDF 다운로드
⟨ 이전 논문 다음 논문 ⟩

초록

스마트 팜 관리의 활용 효율성을 높이기 위해서는 작물 및 환경 변화에 대한 사전 검사를 실시간으로 평가하기 위한 모델링 기법이 필요하다. 시설 온실 내부의 CO2와 같은 필수 환경 요소는 다양한 상관 변수가 밀접하게 결합 된 시간 영역에서 신뢰할 수 있는 추정 모델을 확립하기가 어렵다. 따라서 본 연구는 입력 영역과 출력 변수를 CO2와 같은 시간 관점에서 인접 영역에 분포된 환경 정보를 이용하여 시간 복잡도를 줄이기 위한 인공 신경망을 개발하기 위해 수행되었다. 스마트 팜을 계측하기 위한 센서 모듈을 통해 환경 요소를 지속적으로 측정하였다. 실험기간의 평균 데이터로 예측하는 모델링 1, 전일 데이터로 예측하는 모델링 2을 구성하여 CO2 환경인자의 상호관계를 예측하였다. 전일의 데이터 학습으로 예측하는 모델링 2가 60일 평균값으로 예측한 모델링 1에 비해 성능이 우수하였다. 30일 이전까지는 대부분 0.70~0.88사이의 결정계수를 보였으며 모델링 2가 약0.05정도 높게 나타났다. 하지만 30일 이후에는 두 가지 모델링 모두 결정 계수 값이 0.50 이하로 낮은 값을 보였다. 모델링 접근법에 따라 결정 요인의 값을 비교하고 분석 한 결과 인접한 시간대의 데이터는 고정 신경망 모델을 사용하는 대신 예측이 필요한 지점에서 상대적으로 높은 성능을 나타냈다.

To increase the utilization of the intelligent methodology of smart farm management, estimation modeling techniques are required to assess prior examination of crops and environment changes in realtime. A mandatory environmental factor such as CO2 is challenging to establish a reliable estimation model in time domain accounted for indoor agricultural facilities where various correlated variables are highly coupled. Thus, this study was conducted to develop an artificial neural network for reducing time complexity by using environmental information distributed in adjacent areas from a time perspective as input and output variables as CO2. The environmental factors in the smart farm were continuously measured using measuring devices that integrated sensors through experiments. Modeling 1 predicted by the mean data of the experiment period and modeling 2 predicted by the day-to-day data were constructed to predict the correlation of CO2. Modeling 2 predicted by the previous day's data learning performed better than Modeling 1 predicted by the 60-day average value. Until 30 days, most of them showed a coefficient of determination between 0.70 and 0.88, and Model 2 was about 0.05 higher. However, after 30 days, the modeling coefficients of both models showed low values below 0.50. According to the modeling approach, comparing and analyzing the values of the determinants showed that data from adjacent time zones were relatively high performance at points requiring prediction rather than a fixed neural network model.

키워드

참고문헌

  1. Choi, H.Y., T.W. Moon, D.H. Jung, and J.E. Son. 2019. Prediction of air temperature and relative humidity in greenhouse via a multilayer perceptron using environmental factors. Protected horticulture and plant factory. 28:95-103 (in Korean). https://doi.org/10.12791/KSBEC.2019.28.2.95
  2. Fourati, F. and M. Chtourou. 2007. A greenhouse control with feed-forward and recurrent neural networks. Simulation modelling practice and theory. 15:1016-1028. https://doi.org/10.1016/j.simpat.2007.06.001
  3. Gers, Felix.A., Jrgen. Schmidhuber, and Fred. Cummins. 2000. Neural Computation. 12:2451-2471. https://doi.org/10.1162/089976600300015015
  4. Jo, J.M. 2019. Effectiveness of normalization pre-processing of big data to the machine learning performance. Journal of the KIECS. 14:547-552 (in Korean).
  5. Ju, H.J. 2017. The model evaluation of the greenhouse temperature management and the temperature management scenario to warm root zone during low temperature season for smart farm. Journal of Agriculture & Life Sciences. 48:13-19 (in Korean).
  6. Jung, D.H., H.I. Yoon, and J.E. Son. 2017. Development of a three-variable canopy photosynthetic rate model of romaine lettuce (Lactuca sativa L.) grown in plant factory modules using light intensity, temperature, and growth stage. Protected Horticulture and Plant Factory, 26:268-275 (in Korean). https://doi.org/10.12791/KSBEC.2017.26.4.268
  7. Kim, J.H. 2016. 4th industrial revolution, education in the era of artificial intelligence. STSS sustainability science conference. 21-29 (in Korean).
  8. Kim, S.K., Y.G. Cha, and G.S. Kim. 2016. Smart farm will lead future farming. Samsung KPMG KRI research institute. Issue monitor (in Korean).
  9. Kim, S.K., and T.I. Oh. 2018. Real-time PM10 concentration prediction LSTM model based on IoT streaming sensor data. Korea academy industrial cooperation society. 19:310-318 (in Korean).
  10. Kim, S.Y., K.S.. Park, S.M. Lee, N.M. Heo, and K.H. Ryu. 2018. Development of prediction model for greenhouse control based on machine learning. Journal of digital contents society 19:749-756 (in Korean). https://doi.org/10.9728/DCS.2018.19.4.749
  11. Kwon, H.M., S.E. Hong, D.S. Park, H.J. Kim, and Y.G. Kim. 2017. A study on the analysis of major crop models for optimum growth conditions. Proceedings of the korean institute of communication sciences conference. 1539-1540 (in Korean).
  12. Lee, J.S., Y.G. Hong, K.H. Kim, D.H. Im, and S.R. Han. 2016. A study on the development of cloud system for automatic control of smart greenhouse. Proceedings of the korean institute of communication sciences conference. 559-560 (in Korean).
  13. Ooteghem, R.J.C. 2010. Optimal control design for a solar greenhouse. IFAC proceedings 43:304-309.
  14. Park, S.A., M.G. Kim, M.H. Yoo, M.M. Oh, and K.C. Son. 2010. Comparison of indoor $CO_2$ removal capability of five foliage plants by photosynthesis. Korean journal of horticultural science & technology 28:864-870 (in Korean).
  15. Patil, S.L., H.J. Tantau, and V.M. Salokhe. 2008. Modelling of tropical greenhouse temperature by auto regressive and neural network models. Biosystems engineering 99:423-431. https://doi.org/10.1016/j.biosystemseng.2007.11.009
  16. Statistics Korea. 2017. KOSIS national statistical portal. Available at: http://www.kosis.kr. accessed september 21, 2018.
  17. Trejo-Perea, M., G. Herrera-Ruiz, J. Rios-Moreno, R.C. Miranda and E. Rivas-Araiza. 2009. Greenhouse energy consumption prediction using neural networks models. International journal of agriculture & biology. 235-249.
  18. Vranken, E., R. Gevers, J.M. Aerts and D. Berckmans. 2005. Performance of model-based predictive control of the ventilation rate with axial fans. Biosystems Engineering. 91:87-98. https://doi.org/10.1016/j.biosystemseng.2005.02.006
  19. Zhang, Xiao., Ling. Xu, Feng. Ding, and Tasawar. Hayat. 2018. Combined state and parameter estimation for a bilinear state space system with moving average noise. Journal of the franklin institute. 355:3079-3103. https://doi.org/10.1016/j.jfranklin.2018.01.011