10.1007/s40095-022-00488-3

Near-term, national solar capacity factor forecasts aided by trend attributes and artificial intelligence

  • David A. Wood*1,
  1. DWA Energy Limited, Lincoln, GB

Published in Issue 2022-03-18

How to Cite

Wood, D. A. (2022). Near-term, national solar capacity factor forecasts aided by trend attributes and artificial intelligence. International Journal of Energy and Environmental Engineering, 13(4 (December 2022). https://doi.org/10.1007/s40095-022-00488-3
Crossref
Scopus
Google Scholar
Europe PMC

Abstract

Abstract An attribute technique is applied to forecast countrywide solar capacity. Attributes relate to the prior 12 h of a univariate, hourly time series. The approach avoids uncertainties relating to weather-related variables averaged at the country level. It captures impacts of system curtailments due to abnormal market conditions or grid-offtake limitations. Fifteen attributes relating to each hourly record are input to machine/deep learning (ML/DL) models. 43,824 h of solar capacity factor for Britain from 2015 to 2019 is evaluated. Fifteen ML/DL models are trained with 2015–2018 data with cross-validation. Trained models are then applied to forecast unseen 2019 hourly data. The ML/DL model forecast accuracy is compared with that of ARIMA and regression models. Extreme gradient boosting, random forest and adaptive boosting models outperform ARIMA and regression methods in forecasts for hours t 0 to t  + 12. Those three ML models are more accurate and faster to execute than six DL models evaluated. Suboptimal convergence and/or overfitting hinder the forecasts of DL models with unseen data. A transparent multi-linear regression model is used to identifying attribute influences on the different time period forecasts. The trend attributes are shown to influence the forecasts for different hours ahead in distinct ways.

Keywords

  • Univariate trend attribute analysis,
  • Machine/deep learning,
  • ARIMA,
  • Seasonality factors,
  • Tenfold cross-validation,
  • Solar power time series forecasting

References

  1. Sobri et al. (2018) Solar photovoltaic generation forecasting methods: a review (pp. 459-497) https://doi.org/10.1016/j.enconman.2017.11.019
  2. Badal et al. (2019) A survey on control issues in renewable energy integration and microgrid 4(1) https://doi.org/10.1186/s41601-019-0122-8
  3. Gupta et al. (2021) Countrywide PV hosting capacity and energy storage requirements for distribution networks: the case of Switzerland https://doi.org/10.1016/j.apenergy.2020.116010
  4. Antonanzas et al. (2016) Review of photovoltaic power forecasting (pp. 78-111) https://doi.org/10.1016/j.solener.2016.06.069
  5. Voyant et al. (2017) Machine learning methods for solar radiation forecasting: a review (pp. 569-582) https://doi.org/10.1016/j.renene.2016.12.095
  6. Das et al. (2018) Forecasting of photovoltaic power generation and model optimization: a review (pp. 912-928) https://doi.org/10.1016/j.rser.2017.08.017
  7. Inman et al. (2013) Solar forecasting methods for renewable energy integration 39(6) (pp. 535-576) https://doi.org/10.1016/j.pecs.2013.06.002
  8. Bergin et al. (2017) Large reductions in solar energy production due to dust and particulate air pollution https://doi.org/10.1021/acs.estlett.7b00197
  9. Rana, M., Koprinska, I., Agelidis, V.G.: Solar power forecasting using weather type clustering and ensembles of neural networks. In: International Joint Conference on Neural Networks (IJCNN), Vancouver, BC, Canada, pp. 4962–4969 (2016)
  10. O’Shaughnessya et al. (2020) Too much of a good thing? Global trends in the curtailment of solar PV (pp. 1068-1077) https://doi.org/10.1016/j.solener.2020.08.075
  11. Wood (2019) German solar power generation data mining and prediction with transparent open box learning network integrating weather, environmental and market variables (pp. 354-369) https://doi.org/10.1016/j.enconman.2019.05.114
  12. Alsharif et al. (2019) Time series ARIMA model for prediction of daily and monthly average global solar radiation: the case study of Seoul, South Korea https://doi.org/10.3390/sym11020240
  13. Atique, S., Noureen, S., Roy, V., Subburaj, V., Bayne, S., Macfie, J.: Forecasting of total daily solar energy generation using ARIMA: a case study. In: 2019 IEEE 9th Annual Computing and Communication Workshop and Conference (CCWC), pp. 0114–0119 (2019).
  14. https://doi.org/10.1109/CCWC.2019.8666481
  15. Kushwaha, V., Pindoriya, N.M.: Very short-term solar PV generation forecast using SARIMA model: a case study. In: 7th International Conference on Power Systems, pp. 430–435 (2017).
  16. https://doi.org/10.1109/ICPES.2017.8387332
  17. Soufiane et al. (2020) Solar power time series prediction using wavelet analysis 10(4) (pp. 1764-1773)
  18. Qian et al. (2019) A review and discussion of decomposition-based hybrid models for wind energy forecasting applications 235(1) (pp. 939-953) https://doi.org/10.1016/j.apenergy.2018.10.080
  19. Mandal et al. (2012) Forecasting power output of solar photovoltaic system using wavelet transform and artificial intelligence techniques (pp. 332-337) https://doi.org/10.1016/j.procs.2012.09.080
  20. Kushwaha and Pindoriya (2019) A SARIMA-RVFL hybrid model assisted by wavelet decomposition for very short-term solar PV power generation forecast (pp. 124-139) https://doi.org/10.1016/j.renene.2019.03.020
  21. El Hendouzi and Bourouhou (2020) Solar photovoltaic power forecasting https://doi.org/10.1155/2020/8819925
  22. Ahmed et al. (2020) A review and evaluation of the state-of-the-art in PV solar power forecasting: techniques and optimization https://doi.org/10.1016/j.rser.2020.109792
  23. Golestaneh et al. (2016) Very short-term nonparametric probabilistic forecasting of renewable energy generation—with application to solar energy (pp. 3850-3863) https://doi.org/10.1109/TPWRS.2015.2502423
  24. Abuella, M., Chowdhury, B.: Solar power probabilistic forecasting by using multiple linear regression analysis. In: SoutheastCon 2015, pp. 1–5 (2015).
  25. https://doi.org/10.1109/SECON.2015.7132869
  26. Khatib et al. (2012) Solar energy prediction for Malaysia using artificial neural networks 6(1) (pp. 1-16)
  27. Alfadda, A., Adhikari, R., Kuzlu, M., Rahman, S.: Hour-ahead solar PV power forecasting using SVR based approach. In: 2017 IEEE Power & Energy Society Innovative Smart Grid Technologies Conference (ISGT), Washington, DC, pp. 1–5 (2017)
  28. Meng and Song (2020) Daily photovoltaic power generation forecasting model based on random forest algorithm for north China in winter https://doi.org/10.3390/su12062247
  29. Wang et al. (2018) A short-term photovoltaic power prediction model based on the gradient boost decision tree https://doi.org/10.3390/app8050689
  30. Perveen et al. (2019) An ANFIS-based model for solar energy forecasting and its smart grid application https://doi.org/10.1002/eng2.12070
  31. Ramli, N.A., Hamid, M.F.A., Azhan, N.H., Ishak, M.A.A.: Solar power generation prediction by using
  32. k
  33. -nearest neighbor method. In: AIP Conference Proceedings, p. 020116 (2019).
  34. https://doi.org/10.1063/1.5118124
  35. Wood (2020) Solar plus wind country-wide electrical power forecasts across successive years by optimized data matching (pp. 377-394) https://doi.org/10.1007/s40095-020-00343-3
  36. Gensler, A., Henze, J., Sick, B., Raabe, N.: Deep learning for solar power forecasting—an approach using autoencoder and LSTM neural networks. In: 2016 IEEE International Conference on Systems, Man, and Cybernetics (SMC), pp. 2858–2865 (2016)
  37. Gao et al. (2019) Predicting day-ahead solar irradiance through gated recurrent unit using weather forecasting data https://doi.org/10.1063/1.5110223
  38. Suresh et al. (2020) Forecasting solar PV output using convolutional neural networks with a sliding window algorithm 13(3) https://doi.org/10.3390/en13030723
  39. Alkhayat and Mehmood (2021) A review and taxonomy of wind and solar energy forecasting methods based on deep learning https://doi.org/10.1016/j.egyai.2021.100060
  40. Statsmodel: Statistical models in Python: seasonal decompose (2022).
  41. https://www.statsmodels.org/stable/generated/statsmodels.tsa.seasonal.seasonal_decompose.html
  42. . Accessed 31st Jan 2022
  43. SciKit Learn: Supervised and unsupervised machine learning models in Python (2022).
  44. https://scikit-learn.org/stable/
  45. . Accessed 31st Jan 2022
  46. Huang et al. (2015) Trends in extreme learning machines: a review (pp. 32-48) https://doi.org/10.1016/j.neunet.2014.10.001
  47. Huang et al. (2016) Extreme learning machine for regression and multiclass classification (pp. 513-529) https://doi.org/10.1109/TSMCB.2011.2168604
  48. Subba Rao et al. (2012) An alternative perspective on stochastic coefficient regression models (pp. 445-474) Elsevier https://doi.org/10.1016/B978-0-444-53858-1.00015-6
  49. TensorFlow: Deep learning models in Python (2022).
  50. https://www.tensorflow.org/guide/keras/sequential_model
  51. . Accessed 31st Jan 2022
  52. BayesSearchCV: Bayesian optimization of hyperparameters by SciKit optimization (2022).
  53. https://scikit-optimize.github.io/stable/modules/generated/skopt.BayesSearchCV.html
  54. . Accessed 31st Jan 2022
  55. Sun, W., Zhang, T., Tao, R., Wang, A.: Short-term photovoltaic power prediction modeling based on AdaBoost algorithm and Elman. In: 10th International Conference on Power and Energy Systems, pp. 184–188 (2020).
  56. https://doi.org/10.1109/ICPES51309.2020.9349730
  57. Tan et al. (2017) Ultra-short-term photovoltaic power forecasting in microgrid based on Adaboost clustering 41(21) (pp. 33-39) https://doi.org/10.7500/AEPS20170217006
  58. Li et al. (2020) Photovoltaic power forecasting with a hybrid deep learning approach (pp. 175871-175880) https://doi.org/10.1109/ACCESS.2020.3025860
  59. Detyniecki, M., Marsala, C., Krishnan, A., Siegel, M.: Weather-based solar energy prediction. In: IEEE World Congress on Computational Intelligence June, 10–15 2012 Brisbane, Australia (2012)
  60. Rahul, A., Gupta, A., Bansal, A., Roy, K.: Solar energy prediction using decision tree regressor. In: 5th International Conference on Intelligent Computing and Control Systems, pp. 489–495 (2021).
  61. https://doi.org/10.1109/ICICCS51141.2021.9432322
  62. Jing, B., Qian, Z., Pei, Y., Wang, J.: Ultra-short-term PV power forecasting based on ELM segmentation mode. In: The 6th International Conference on Renewable Power Generation Journal of Engineering, vol. 13, pp. 2564–2568 (2017).
  63. https://doi.org/10.1049/joe.2017.0790
  64. Sahu et al. (2021) Short/medium term solar power forecasting of Chhattisgarh state of India using modified TLBO optimized ELM https://doi.org/10.1016/j.jestch.2021.02.016
  65. Dairi et al. (2020) Short-term forecasting of photovoltaic solar power production using variational auto-encoder driven deep learning approach https://doi.org/10.3390/app10238400
  66. Wojtkiewicz et al. (2019) Hour-ahead solar irradiance forecasting using multivariate gated recurrent units https://doi.org/10.3390/en12214055
  67. Hosseini et al. (2020) Direct normal irradiance forecasting using multivariate gated recurrent units https://doi.org/10.3390/en13153914
  68. Chen and Kartini (2017) K-nearest neighbor neural network models for very short-term global solar irradiance forecasting based on meteorological data 10(2) https://doi.org/10.3390/en10020186
  69. Qing and Niu (2018) Hourly day-ahead solar irradiance prediction using weather forecasts by LSTM (pp. 461-468) https://doi.org/10.1016/j.energy.2018.01.177
  70. Abdel-Nasser and Mahmoud (2019) Accurate photovoltaic power forecasting models using deep LSTM-RNN (pp. 2727-2740) https://doi.org/10.1007/s00521-017-3225-z
  71. Kong et al. (2019) Short term residential load forecasting based on LSTM recurrent neural network 10(1) (pp. 841-851) https://doi.org/10.1109/TSG.2017.2753802
  72. Ehsan et al. (2017) Day-ahead forecasting of solar photovoltaic output power using multilayer perceptron (pp. 3981-3992) https://doi.org/10.1007/s00521-016-2310-z
  73. Parvez, I., Sarwat, A., Debnath, A., Olowu, T., Dastgir, M.G., Riggs, H.: Multi-layer perceptron based photovoltaic forecasting for rooftop PV applications in smart grid. In: 2020 SoutheastCon, pp. 1–6 (2020).
  74. https://doi.org/10.1109/SoutheastCon44009.2020.9249681
  75. Yadav et al. (2020) ANN and multiple linear regression based modelling for experimental investigation of photovoltaic module maximum power production under outdoor condition of mountainous region Springer https://doi.org/10.1007/978-3-030-05578-3_8
  76. Kim et al. (2021) Use of a big data analysis in regression of solar power generation on meteorological variables for a Korean solar power plant https://doi.org/10.3390/app11041776
  77. Tato and Brito (2019) Using smart persistence and random forests to predict photovoltaic energy production https://doi.org/10.3390/en12010100
  78. Chen et al. (2011) Estimation of monthly solar radiation from measured temperatures using support vector machines 36(1) (pp. 413-420) https://doi.org/10.1016/j.renene.2010.06.024
  79. Fentis, A., Bahatti, L., Mestari, M., Chouri, B.: Short-term solar power forecasting using support vector regression and feed-forward NN. In: 15th IEEE International New Circuits and Systems Conference, pp. 405–408 (2017).
  80. https://doi.org/10.1109/NEWCAS.2017.8010191
  81. Buwei, W., Jianfeng, C., Bo, W., Shuanglei, F.: A solar power prediction using support vector machines based on multi-source data fusion. In: 2018 International Conference on Power System Technology, pp. 4573–4577 (2018).
  82. https://doi.org/10.1109/POWERCON.2018.8601672
  83. Aguilar Madrid and Antonio (2021) Short-term electricity load forecasting with machine learning https://doi.org/10.3390/info12020050
  84. NASA: MERRA-2 database from the National Aeronautics and Space Administration satellites (2022).
  85. www.renewables.ninja
  86. and
  87. https://gmao.gsfc.nasa.gov/reanalysis/MERRA-2/
  88. . Accessed 2nd Feb 2022
  89. Fara et al. (2021) Forecasting of energy production for photovoltaic systems based on ARIMA and ANN advanced models https://doi.org/10.1155/2021/6777488
  90. Mbaye, A., Ndiaye, M., Ndione, D., Diaw, M., Traoré, V., et al.: ARMA model for short-term forecasting of solar potential: application to a horizontal surface on Dakar site. Collaborating Academics–International Press 4(1), (2019).
  91. https://doi.org/10.23647/ca.md20191103
  92. . hal-02151290.
  93. https://hal.archives-ouvertes.fr/hal-02151290