The Korean Journal of Applied Statistics (응용통계연구)

The Korean Statistical Society (한국통계학회)
권호 표지
DOI QR코드

DOI QR Code

Solar radiation forecasting by time series models

시계열 모형을 활용한 일사량 예측 연구

  • Suh, Yu Min (Department of Applied Statistics, Chung-Ang University) ;
  • Son, Heung-goo (Korea Power Exchange) ;
  • Kim, Sahm (Department of Applied Statistics, Chung-Ang University)
  • Received : 2018.10.25
  • Accepted : 2018.12.11
  • Published : 2018.12.31
Download PDF
⟨ Previous Next ⟩

Abstract

With the development of renewable energy sector, the importance of solar energy is continuously increasing. Solar radiation forecasting is essential to accurately solar power generation forecasting. In this paper, we used time series models (ARIMA, ARIMAX, seasonal ARIMA, seasonal ARIMAX, ARIMA GARCH, ARIMAX-GARCH, seasonal ARIMA-GARCH, seasonal ARIMAX-GARCH). We compared the performance of the models using mean absolute error and root mean square error. According to the performance of the models without exogenous variables, the Seasonal ARIMA-GARCH model showed better performance model considering the problem of heteroscedasticity. However, when the exogenous variables were considered, the ARIMAX model showed the best forecasting accuracy.

신재생에너지 산업이 발전함에 따라 태양광 발전에 대한 중요성이 확대되고 있다. 태양광 발전량을 정확히 예측하기 위해서는 일사량 예측이 필수적이다. 본 논문에서는 태양광 패널이 존재하는 청주와 광주 지역을 선정하여 기상포털에서 제공하는 시간별 기상 데이터를 수집하여 연구하였다. 일사량 예측을 위하여 시계열 모형인 ARIMA, ARIMAX, seasonal ARIMA, seasonal ARIMAX, ARIMA-GARCH, ARIMAX-GARCH, seasonal ARIMA-GARCH, seasonal ARIMAX-GARCH 모형을 비교하였다. 본 연구에서는 모형의 예측 성능을 비교하고자 mean absolute error와 root mean square error를 사용하였다. 모형들의 예측 성능 비교 결과 일사량만 고려하였을 때는 이분산 문제를 고려한 seasonal ARIMA-GARCH 모형이 우수한 성능을 나타냈고, 외생변수를 활용한 ARIMAX 모형으로 일사량 예측을 한 경우가 가장 좋은 예측력을 나타냈다.

Keywords

GCGHDE_2018_v31n6_785_f0001.png 이미지

Figure 3.1. Hourly solar radiation of Cheongju and Gwangju in 2015.

GCGHDE_2018_v31n6_785_f0002.png 이미지

Figure 3.2. Hourly solar radiation of Cheongju and Gwangju in January 2015.

GCGHDE_2018_v31n6_785_f0003.png 이미지

Figure 3.3. Result of forecasting 7-Days ahead, ARIMA model.

GCGHDE_2018_v31n6_785_f0004.png 이미지

Figure 3.4. Correlation of Cheongju and Gwangju.

GCGHDE_2018_v31n6_785_f0005.png 이미지

Figure 3.5. Result of forecasting 7-Days ahead, ARIMAX model. ARIMAX = auto-regressive integrated moving average with eXogenous variable.

GCGHDE_2018_v31n6_785_f0006.png 이미지

Figure 3.6. Result of Forecasting 7-Days ahead, seasonal ARIMA model. ARIMA = auto-regressive integratedmoving average.

GCGHDE_2018_v31n6_785_f0007.png 이미지

Figure 3.7. Result of forecasting 7-Days ahead, seasonal ARIMAX model. ARIMAX = auto-regressive integrated moving average with eXogenous variable.

GCGHDE_2018_v31n6_785_f0008.png 이미지

Figure 3.8. Result of forecasting 7-Days ahead, ARIMA-GARCH model. ARIMA = auto-regressive integrated moving average; GARCH = generalized auto-regressive conditionally heteroscadastic.

GCGHDE_2018_v31n6_785_f0009.png 이미지

Figure 3.9. Result of forecasting 7-Days ahead, ARIMAX-GARCH model. ARIMAX = auto-regressive integrated moving average with eXogenous variable; GARCH = generalized auto-regressive conditionally heteroscadastic.

GCGHDE_2018_v31n6_785_f0010.png 이미지

Figure 3.10. Result of forecasting 7-Days ahead, seasonal ARIMA-GARCH model. ARIMA = auto-regressive integrated moving average; GARCH = generalized auto-regressive conditionally heteroscadastic.

GCGHDE_2018_v31n6_785_f0011.png 이미지

Figure 3.11. Result of forecasting 7-Days ahead, seasonal ARIMAX-GARCH model. ARIMAX = auto-regressive integrated moving average with eXogenous variable; GARCH = generalized auto-regressive conditionally heteroscadastic.

Table 3.1. Fitted models using ARIMA

GCGHDE_2018_v31n6_785_t0001.png 이미지

Table 3.2. Result of variable selection using ARIMAX model

GCGHDE_2018_v31n6_785_t0002.png 이미지

Table 3.3. Fitted models using ARIMAX

GCGHDE_2018_v31n6_785_t0003.png 이미지

Table 3.4. Fitted models using seasonal ARIMA

GCGHDE_2018_v31n6_785_t0004.png 이미지

Table 3.5. Fitted models using seasonal ARIMAX

GCGHDE_2018_v31n6_785_t0005.png 이미지

Table 3.6. Model comparison

GCGHDE_2018_v31n6_785_t0006.png 이미지

Table 3.7. Model comparison of peak hour using mean absolute percentage error

GCGHDE_2018_v31n6_785_t0007.png 이미지

References

  1. Colak, I., Yesilbudak, M., Genc, N., and Bayindir, R. (2015). Multi-period prediction of solar radiation using ARMA and ARIMA models, In Machine Learning and Applications (ICMLA), 2015 IEEE 14th International Conference on (pp. 1045-1049). IEEE.
  2. Elizondo, D., Hoogenboom, G., and McClendon, R. W. (1994). Development of a neural network model to predict daily solar radiation, Agricultural and Forest Meteorology, 71, 115-132. https://doi.org/10.1016/0168-1923(94)90103-1
  3. Grek, A. (2014). Forecasting accuracy for ARCH models and GARCH (1, 1) family: Which model does best capture the volatility of the Swedish stock market?.
  4. Ji, W. and Chee, K. C. (2011). Prediction of hourly solar radiation using a novel hybrid model of ARMA and TDNN, Solar Energy, 85, 808-817. https://doi.org/10.1016/j.solener.2011.01.013
  5. Kim, S. (2017). A study on solar irradiance forecasting with weather variables. The Korean Journal of Applied Statistics, 30(6), 1005-1013. https://doi.org/10.5351/KJAS.2017.30.6.1005
  6. Kim, S., Lee, K., and Son, H. (2018). A study on solar energy forecasting based on time series models. The Korean Journal of Applied Statistics, 31(1), 139-153. https://doi.org/10.5351/KJAS.2018.31.1.139
  7. Kim, S. and Park S. (2016). A study on short-term wind power forecasting using time series models, The Korean Journal of Applied Statistics, 29, 1373-1383.
  8. Liu, H., Erdem, E., and Shi, J. (2011). Comprehensive evaluation of ARMA-GARCH (-M) approaches for modeling the mean and volatility of wind speed. Applied Energy, 88, 724-732. https://doi.org/10.1016/j.apenergy.2010.09.028
  9. Rehman, S. and Mohandes, M. (2008). Artificial neural network estimation of global solar radiation using air temperature and relative humidity. Energy Policy, 36, 571-576. https://doi.org/10.1016/j.enpol.2007.09.033
  10. Sun, H., Yan, D., Zhao, N., and Zhou, J. (2015). Empirical investigation on modeling solar radiation series with ARMA-GARCH models. Energy Conversion and Management, 92, 385-395. https://doi.org/10.1016/j.enconman.2014.12.072