The Journal of the Korea institute of electronic communication sciences (한국전자통신학회논문지)

Korea Institute of Electronic Communication Science (한국전자통신학회)
권호 표지
DOI QR코드

DOI QR Code

Very Short- and Long-Term Prediction Method for Solar Power

초 장단기 통합 태양광 발전량 예측 기법

  • 윤문섭 (전남대학교 전자통신공학과) ;
  • 임세령 (전남대학교 전자통신공학과) ;
  • 장한승 (전남대학교 전자통신공학과)
  • Received : 2023.10.06
  • Accepted : 2023.12.27
  • Published : 2023.12.31
Download PDF
⟨ Previous Next ⟩

Abstract

The global climate crisis and the implementation of low-carbon policies have led to a growing interest in renewable energy and a growing number of related industries. Among them, solar power is attracting attention as a representative eco-friendly energy that does not deplete and does not emit pollutants or greenhouse gases. As a result, the supplement of solar power facility is increasing all over the world. However, solar power is easily affected by the environment such as geography and weather, so accurate solar power forecast is important for stable operation and efficient management. However, it is very hard to predict the exact amount of solar power using statistical methods. In addition, the conventional prediction methods have focused on only short- or long-term prediction, which causes to take long time to obtain various prediction models with different prediction horizons. Therefore, this study utilizes a many-to-many structure of a recurrent neural network (RNN) to integrate short-term and long-term predictions of solar power generation. We compare various RNN-based very short- and long-term prediction methods for solar power in terms of MSE and R2 values.

세계적 기후 위기와 저탄소 정책 이행으로 신재생 에너지에 관한 관심이 높아지고 이와 관련된 산업이 증가하고 있다. 이 중에서 태양 에너지는 고갈되지 않고 오염 물질이나 온실가스를 배출하지 않는 대표적인 친환경 에너지로 주목받고 있으며, 이에 따라 세계적으로 태양광 발전 시설 보급이 증가하고 있다. 하지만 태양광 발전은 지리, 날씨와 같은 환경의 영향을 받기 쉬우므로 안정적인 운영과 효율적인 관리를 위해 정확한 발전량 예측이 중요하다. 하지만 변동성이 큰 태양광 발전을 수학적 통계 기술로 정확한 발전량을 예측하는 것은 불가능하다. 이를 위해서 정확하고 효과적인 예측을 위해 딥러닝 기반의 기술에 관한 연구는 필수적이다. 또한, 기존의 딥러닝을 활용한 예측 방식은 장, 단기적인 예측을 나누어 수행하기 때문에 각각의 예측 결과를 얻기 위한 시간이 길어진다는 단점이 있다. 따라서, 본 연구에서는 시계열 특성을 가진 태양광 발전량 데이터를 사용하여 장단기 통합 예측을 수행하기 위해 순환 신경망의 다대다 구조를 활용한다. 그리고 이를 다양한 딥러닝 모델들에 적용하여 학습을 수행하고 각 모델의 결과를 비교·분석한다.

Keywords

Acknowledgement

본 논문은 전남대학교 학술연구비(과제번호: 2021-2176) 지원에 의하여 연구되었음.

References

  1. H. Kim, G. Yang, C. Nam, S. Jeong and S. P. Jung,"Solar photovoltaic industry in Korea: cur rent status and perspectives,"J. of Korean Society of Environmental Engineers, vol. 45, no. 2, Fe b. 2023, pp. 107-119. https://doi.org/10.4491/KSEE.2023.45.2.107
  2. M. Kim, S. Jung, J. Kim, H. Lee, B. Kim and S. Kim,"A Study on Artificial Neural Network-based Solar Radiation Forecasting for Efficient Solar Photovoltaic System,"J. of Korean Institute of Intelligent Systems, vol. 29, no. 6, Dec. 2019, pp. 501-506. https://doi.org/10.5391/JKIIS.2019.29.6.501
  3. S.-M. Lee, J.-A. Noh, S.-J. Kang and J.-W. Park,"A Study on Machine Learning Models and Trends for Forecasting of Photovoltaic Power Generation,"Conf. The Korean Institute of Electrical Engineers, Oct. 2021, pp. 169-170.
  4. S.-R. Jung, J. Koh and S.-K. Lee, "Recurrent Network based Prediction System of Agricultural Photovoltaic Power Generation," J. of the Korea institute of Electronic Communication, vol. 17, no. 5, 2022, pp. 825-832.
  5. M. A. Bou-Rabee, M. Y. Naz, I. E. Albalaa and S. A. Sulaiman,"BiLSTM Network-Based Approach for Solar Irradiance Forecasting in Continental Climate Zones," Energies, vol. 15, no. 6, 2022, 2226.
  6. S.-R. Jung, K.-W. Park and S.-K. Lee, "Intelligent Prediction System for Diagnosis of Agricultural Photovoltaic Power Generation," J. of the Korea institute of Electronic Communication, vol. 16, no. 5, 2021, pp. 859-866.
  7. M.-Y. Kang, "Renewable Energy Generation Prediction Model using Meteorological Big Data,"J. of the Korea institute of Electronic Communication, vol. 18, no. 1, 2023, pp. 39-44.
  8. Y. Zhou, Y. Li, D. wang, and Y. Liu"A multi-step ahead global solar radiation prediction method using an attention-based transformer model with an interpretable mechanism,"Int. J. of Hydrogen Energy, vol. 48, no. 40, May, 2023, pp. 15317-15330.
  9. A. Heidari and D. Khovalyg,"Short-term energy use prediction of solar-assisted water heating system: Application case of combined attention-based LSTM and time-series decomposition,"Solar Energy vol. 207, no 1, Sept., 2020, pp. 626-639. https://doi.org/10.1016/j.solener.2020.07.008
  10. A. Graves, "Long short-term memory,"Supervised sequence labelling with recurrent neural networks, Berlin : Springer, 2012, pp. 37-45.
  11. T. Peng, C. Zhang, J. Zhou and M. S. Nazir, "An integrated framework of Bi-directional lo ng-short term memory (BiLSTM) based on sine cosine algorithm for hourly solar radiation forecasting,"Energy vol. 221, Apr., 2021. pp. 119887.
  12. A. Vaswani, N. Shazeer, N. Parmar, J. Uszkoreit, L. Jones, A. N. Gomesz and I. Polosukhin,"Attention is all you need," Advances in neural information processing systems 30, 2017.
  13. N. Rahimi, S. Park, W. Choi, B. Oh, S. Kim, Y. Cho and D. Lee, "A Comprehensive Review on Ensemble Solar Power Forecasting Algorithms."J. of Electrical Engineering & Technology, Jan., 2023, pp. 719-733.