Journal of Advanced Navigation Technology (한국항행학회논문지)

The Korean Navigation Institute (한국항행학회)
권호 표지
DOI QR코드

DOI QR Code

Short Term Forecast Model for Solar Power Generation using RNN-LSTM

RNN-LSTM을 이용한 태양광 발전량 단기 예측 모델

  • 신동하 (가천대학교 에너지 IT학과) ;
  • 김창복 (가천대학교 에너지 IT학과)
  • Received : 2018.04.23
  • Accepted : 2018.06.19
  • Published : 2018.06.30
Download PDF
⟨ Previous Next ⟩

Abstract

Since solar power generation is intermittent depending on weather conditions, it is necessary to predict the accurate generation amount of solar power to improve the efficiency and economical efficiency of solar power generation. This study proposes a short - term deep learning prediction model of solar power generation using meteorological data from Mokpo meteorological agency and generation data of Yeongam solar power plant. The meteorological agency forecasts weather factors such as temperature, precipitation, wind direction, wind speed, humidity, and cloudiness for three days. However, sunshine and solar radiation, the most important meteorological factors for forecasting solar power generation, are not predicted. The proposed model predicts solar radiation and solar radiation using forecast meteorological factors. The power generation was also forecasted by adding the forecasted solar and solar factors to the meteorological factors. The forecasted power generation of the proposed model is that the average RMSE and MAE of DNN are 0.177 and 0.095, and RNN is 0.116 and 0.067. Also, LSTM is the best result of 0.100 and 0.054. It is expected that this study will lead to better prediction results by combining various input.

태양광 발전은 기상 상태에 따라 간헐적이기 때문에 태양광 발전의 효율과 경제성 향상을 위해 정확한 발전량 예측이 요구된다. 본 연구는 목포 기상대에서 예보하는 기상 데이터와 영암 태양광 발전소의 발전량 데이터를 이용하여 태양광 발전량 단기 딥러닝 예측모델을 제안하였다. 기상청은 기온, 강수량, 풍향, 풍속, 습도, 운량 등의 기상요소를 3일간 예보한다. 그러나 태양광 발전량 예측에 가장 중요한 기상요소인 일조 및 일사 일사량 예보하지 않는다. 제안 모델은 예보 기상요소를 이용하여, 일조 및 일사 일사량을 예측 하였다. 또한 발전량은 기상요소에 예측된 일조 및 일사 기상요소를 추가하여 예측하였다. 제안 모델의 발전량 예측 결과 DNN의 평균 RMSE와 MAE는 0.177과 0.095이며, RNN은 0.116과 0.067이다. 또한, LSTM은 가장 좋은 결과인 0.100과 0.054이다. 향후 본 연구는 다양한 입력요소의 결합으로 보다 향상된 예측결과를 도출할 수 있을 것으로 기대된다.

Keywords

References

  1. S. M. Lee and Y. H. Chun, “Assessment of optimal constitution rate of wind turbine and photovoltaic sources for stable operation of microgird,” The Transactions of The Korean Institute of Electrical Engineers, Vol. 59, No. 2, pp. 272-276, Feb. 2010
  2. B. H. Lee, “A study on simplified robust optimal operation of microgrids considering the uncertainty of renewable generation and loads,” The Transactions of The Korean Institute of Electrical Engineers, Vol. 66, No. 3, pp. 513-521, May. 2017 https://doi.org/10.5370/KIEE.2017.66.3.513
  3. S. B. Rhee, K. H. Kim, and S. G. Lee, "Optimal operation scheme of microgrid system based on renewable energy resources," The Transactions of the Korean Institute of Electrical Engineers, Vol. 60, No. 8, pp. 1467-1472, Aug. 2011 https://doi.org/10.5370/KIEE.2011.60.8.1467
  4. M. H. Seo, G. S. Kim, and S. H. Kim, "A development of the solar position algorithm for improving the efficiency of photovoltaic power generation," in Proceedings of KIIT Summer Conference, Vol. 8, No. 10, Mokpo: Korea, pp. 46-51, June. 2009
  5. J. J. Song, Y. S. Jeong, and S. H. Lee, “Analysis of prediction model for solar power generation,” Journal of Digital Convergence, Vol. 12, No. 3, pp. 243-248, Mar. 2014 https://doi.org/10.14400/JDC.2014.12.3.243
  6. K. D. Kim, “The development of the short-term predict model for solar power generation,” The Korea Solar Energy Society, Vol. 33, No. 6, pp. 62-69, Dec. 2013 https://doi.org/10.7836/kses.2013.33.6.062
  7. C. S. Lee and P. S. Ji, “Development of daily PV power forecasting models using ELM,” The Transactions of The Korean Institute of Electrical Engineers, Vol. 64P, No. 3, pp. 164-168, Sep. 2015
  8. K. H. Lee and W. J. Kim, “Forecasting of 24 hours ahead photovoltaic power output using support vector regression,” Journal of Korean Institute of Information Technology, Vol. 14, No. 3, pp. 175-183, May. 2016
  9. D. J. Lee, J. P. Lee, C. S. Lee, J. Y. Lim, and P. S. Ji, “Development of PV power prediction algorithm using adaptive neuro-fuzzy model,” The Transactions of The Korean Institute of Electrical Engineers, Vol. 64, No. 4, pp. 246-250, Dec. 2015 https://doi.org/10.5370/KIEEP.2015.64.4.246
  10. W. C. Cha, J. H. Park, U. R. Cho, and J. C. Kim, "Design of generation efficiency fuzzy prediction model using solar power element data," The Transactions of The Korean Institute of Electrical Engineers, Vol. 63, No. 10, pp. 1423-1427, Oct. 2014 https://doi.org/10.5370/KIEE.2014.63.10.1423
  11. S. M. Lee and W. J. Lee, “Development of a system for predicting photovoltaic power generation and detecting defects using machine learning,” KIPS Transactions on Computer and Communication Systems, Vol. 5, No. 10, pp. 353-360, Oct. 2016 https://doi.org/10.3745/KTCCS.2016.5.10.353
  12. A. Yona, T. Senjyu, T. Funabashi, P. Mandal, and C. H. Kim, “Decision technique of solar radiation prediction applying recurrent neural network for short-term ahead power output of photovoltaic system,” Smart Grid and Renewable Energy, Vol. 4, No. 6A, pp. 32-38, Apr. 2013 https://doi.org/10.4236/sgre.2013.46A004
  13. F. A. Gers, N. N. Schraudolph, and J. Schmidhuber, "Learning precise timing with LSTM recurrent networks," Journal of Machine Learning Research, Vol. 3, No. 6, pp. 115-143, Mar. 2002
  14. C. Olah, Understanding LSTM Networks, Github blog [Internet]. Available: http://colah.github.io/posts/2015-08-Understanding-LSTMs/.

Cited by

  1. Predictive model for PV power generation using RNN (LSTM) vol.35, pp.2, 2018, https://doi.org/10.1007/s12206-021-0140-0