The Korean Journal of Applied Statistics (응용통계연구)

The Korean Statistical Society (한국통계학회)
권호 표지
DOI QR코드

DOI QR Code

A study on solar radiation prediction using medium-range weather forecasts

중기예보를 이용한 태양광 일사량 예측 연구

  • Sujin Park (Department of Applied Statistics, Chung-Ang University) ;
  • Hyojeoung Kim (Department of Applied Statistics, Chung-Ang University) ;
  • Sahm Kim (Department of Applied Statistics, Chung-Ang University)
  • 박수진 (중앙대학교 응용통계학과) ;
  • 김효정 (중앙대학교 응용통계학과) ;
  • 김삼용 (중앙대학교 응용통계학과)
  • Received : 2022.10.13
  • Accepted : 2022.11.09
  • Published : 2023.02.28
Download PDF
⟨ Previous Next ⟩

Abstract

Solar energy, which is rapidly increasing in proportion, is being continuously developed and invested. As the installation of new and renewable energy policy green new deal and home solar panels increases, the supply of solar energy in Korea is gradually expanding, and research on accurate demand prediction of power generation is actively underway. In addition, the importance of solar radiation prediction was identified in that solar radiation prediction is acting as a factor that most influences power generation demand prediction. In addition, this study can confirm the biggest difference in that it attempted to predict solar radiation using medium-term forecast weather data not used in previous studies. In this paper, we combined the multi-linear regression model, KNN, random fores, and SVR model and the clustering technique, K-means, to predict solar radiation by hour, by calculating the probability density function for each cluster. Before using medium-term forecast data, mean absolute error (MAE) and root mean squared error (RMSE) were used as indicators to compare model prediction results. The data were converted into daily data according to the medium-term forecast data format from March 1, 2017 to February 28, 2022. As a result of comparing the predictive performance of the model, the method showed the best performance by predicting daily solar radiation with random forest, classifying dates with similar climate factors, and calculating the probability density function of solar radiation by cluster. In addition, when the prediction results were checked after fitting the model to the medium-term forecast data using this methodology, it was confirmed that the prediction error increased by date. This seems to be due to a prediction error in the mid-term forecast weather data. In future studies, among the weather factors that can be used in the mid-term forecast data, studies that add exogenous variables such as precipitation or apply time series clustering techniques should be conducted.

급속적으로 비중이 증가하고 있는 태양광 에너지는 지속적인 개발 및 투자가 이루어지고 있다. 신재생에너지 정책인 그린뉴딜과 가정용 태양광 패널의 설치가 증가함에 따라 국내 태양광 에너지 보급이 점차 확대되어 그에 맞추어 발전량의 정확한 수요 예측 연구가 활발하게 진행되고 있는 시점이다. 또한, 일사량 예측이 발전량 수요 예측에 가장 영향을 미치는 요소로 작용하고 있다는 점에서 일사량 예측의 중요성을 파악하였다. 덧붙여, 본 연구는 선행 연구들에서 사용되지 않은 중기예보 기상 데이터를 활용하여 일사량 예측을 하고자 하였다는 점에서 가장 큰 차이점을 확인할 수 있다. 본 논문에서는 서울, 인천, 수원, 춘천, 대구, 대전의 총 여섯 지역의 태양광 일사량 예측을 위하여 다중선형회귀모형, KNN, Random Forest 그리고 SVR 모형과 클러스터링 기법인 K-means 기법을 결합한 후, 클러스터별 확률밀도함수를 계산하여 시간별 일사량 예측을 진행하고자 하였다. 중기예보 데이터를 사용하기 전, 모형 예측 결과를 비교하기 위한 지표로서 MAE (mean absolute error)와 RMSE (root mean squared error)를 사용하였다. 데이터는 2017년 3월 1일부터 2022년 2월 28일까지의 시간별 원 관측 데이터를 중기예보 데이터 양식에 맞추어 일별 데이터로 변환하였다. 모형의 예측 성능 비교 결과, Random Forest로 일별 일사량을 예측한 후, K-means 클러스터링으로 기후요인이 유사한 날짜들을 분류한 뒤 클러스터별 일사량의 확률밀도함수를 계산하여 시간별 일사량 예측값을 나타낸 방법이 가장 우수한 성능을 보였다. 또한 이 방법론을 이용하여 중기예보 데이터에 모형 적합 후, 예측 결과를 확인하였을 때, 일자별로 예측 오류가 상승하는 것을 확인할 수 있었다. 이는 중기예보 기상데이터의 예측 오류로 인한 것으로 보인다. 향후 연구에서는 중기예보 데이터에서 활용할 수 있는 기상요인 중, 강수 여부와 같은 외생 변수를 추가하거나 시계열 클러스터링 기법을 적용한 연구가 이루어져야할 것으로 보인다.

Keywords

Acknowledgement

이 논문은 2022년도 중앙대학교 CAU GRS 지원에 의하여 작성되었음.

References

  1. Abuella M and Chowdhury B (2015). Solar power probabilistic forecasting by using multiple linear regression analysis. In Proceedings of SoutheastCon 2015, Fort Lauderdale, FL, 1-5.
  2. Alzahrani A, Shamsi P, Dagli C, and Ferdowsi M (2017). Solar irradiance forecasting using deep neural networks, Procedia Computer Science, 114, 304-313. https://doi.org/10.1016/j.procs.2017.09.045
  3. Al-Hajj R, Assi A, and Fouad MM (2019). Stacking-based ensemble of support vector regressors for one-day ahead solar irradiance prediction. In Proceedings of 2019 8th IEEE International Conference on Renewable Energy Research and Applications (ICRERA), 428-433.
  4. Bamisile O, Oluwasanmi A, Ejiyi C, Yimen N, Obiora S, and Huang Q (2022). Comparison of machine learning and deep learning algorithms for hourly global/diffuse solar radiation predictions, International Journal of Energy Research, 46, 10052-10073. https://doi.org/10.1002/er.6529
  5. Benmouiza K and Cheknane A (2013). Forecasting hourly global solar radiation using hybrid k-means and nonlinear autoregressive neural network models, Energy Conversion and Management, 75, 561-569. https://doi.org/10.1016/j.enconman.2013.07.003
  6. Breiman L (2001). Random forests, Machine learning, 45, 5-32. https://doi.org/10.1023/A:1010933404324
  7. Demir V and Citakoglu H (2022). Forecasting of solar radiation using different machine learning approaches, Neural Computing and Applications, 35, 887-906. https://doi.org/10.1007/s00521-022-07841-x
  8. Fan GF, Guo YH, Zheng JM, and Hong WC (2019). Application of the weighted k-nearest neighbor algorithm for short-term load forecasting, Energies, 12, 916.
  9. Fan J, Wang X, Zhang F, Ma X, and Wu L (2020). Predicting daily diffuse horizontal solar radiation in various climatic regions of China using support vector machine and tree-based soft computing models with local and extrinsic climatic data, Journal of Cleaner Production, 248, 119264.
  10. Gala Y, Fernandez A, Diaz J, and Dorronsoro JR (2013). Support vector forecasting of solar radiation values. In International Conference on Hybrid Artificial Intelligence Systems : Vol. 8073. LNAI (pp. 51-60), Springer, Berlin.
  11. Ghofrani M, Ghayekhloo M, and Azimi R (2016). A novel soft computing framework for solar radiation forecasting, Applied Soft Computing, 48, 207-216. https://doi.org/10.1016/j.asoc.2016.07.022
  12. Hong T (2010). Short Term Electric Load Forecasting, North Carolina State University, Raleigh.
  13. Jawaid F and NazirJunejo K (2016). Predicting daily mean solar power using machine learning regression techniques. In Proceedings of 2016 6th IEEE International Conference on Innovative Computing Technology (INTECH), 355-360.
  14. Jimenez-Perez PF and Mora-Lopez L (2016). Modeling and forecasting hourly global solar radiation using clustering and classification techniques, Solar Energy, 135, 682-691. https://doi.org/10.1016/j.solener.2016.06.039
  15. Kim H, Park S, and Kim S (2022). Solar radiation forecasting using boosting decision tree and recurrent neural networks, Communications for Statistical Applications and Methods, 29, 709-719. https://doi.org/10.29220/CSAM.2022.29.6.709
  16. Karasu S and Altan A (2019). Recognition model for solar radiation time series based on random forest with feature selection approach. In Proceedings of 2019 11th IEEE International Conference on Electrical and Electronics Engineering (ELECO), 8-11.
  17. Khosravi A, Koury RNN, Machado L, and Pabon JJG (2018). Prediction of hourly solar radiation in Abu Musa Island using machine learning algorithms, Journal of Cleaner Production, 176, 63-75. https://doi.org/10.1016/j.jclepro.2017.12.065
  18. Ozturk I (2013). Energy dependency and energy security: The role of energy efficiency and renewable energy sources, The Pakistan Development Review, 52, 309-330. https://doi.org/10.30541/v52i4Ipp.309-330
  19. Park S, Lee JY, and Kim S (2021). Wind power forecasting based on time series and machine learning models, The Korean Journal of Applied Statistics, 34, 723-734. https://doi.org/10.5351/KJAS.2021.34.5.723
  20. Rousseeuw PJ (1987). Silhouettes: a graphical aid to the interpretation and validation of cluster analysis, Journal of Computational and Applied Mathematics, 20, 53-65. https://doi.org/10.1016/0377-0427(87)90125-7
  21. Srivastava R, Tiwari AN, and Giri VK (2019). Solar radiation forecasting using MARS, CART, M5, and random forest model: A case study for India, Heliyon, 5, e02692.
  22. Tambunan HB, Barus DH, Hartono J, Alam AS, Nugraha DA, and Usman HHH (2020). Electrical peak load clustering analysis using K-means algorithm and silhouette coefficient. In Proceedings of 2020 International Conference on Technology and Policy in Energy and Electric Power (ICT-PEP), IEEE, Bandung, 258-262.
  23. Tranmer M and Elliot M (2008). Multiple linear regression, The Cathie Marsh Centre for Census and Survey Research (CCSR), 5, 1-5.
  24. Vapnik V (1999). The Nature of Statistical Learning Theory, Springer science & business media, NewYork.
  25. Yadav AP and Behera L (2014). Solar Radiation forecasting using neural networks and Wavelet Transform, IFAC Proceedings Volumes, 47, 890-896. https://doi.org/10.3182/20140313-3-IN-3024.00218
  26. Zhang Z, Wang C, Peng X, Qin H, Lv H, Fu J, and Wang H (2021). Solar radiation intensity probabilistic forecasting based on K-means time series clustering and Gaussian process regression, IEEE Access, 9, 89079-89092. https://doi.org/10.1109/ACCESS.2021.3077475