Pengembangan Model Prediksi Daya Surya Jangka Pendek Menggunakan Ensemble Random Forest dengan Rekayasa Fitur Temporal

Christio Revano Mege; Ferizandi Qauzar Gani; Septia Eka Marsha Putra; Friska Hasugian

doi:10.56211/sudo.v5i3.1998

Authors

Christio Revano Mege Institut Teknologi Sumatera
Ferizandi Qauzar Gani Institut Teknologi Sumatera
Septia Eka Marsha Putra Institut Teknologi Sumatera
Friska Hasugian Institut Teknologi Sumatera

DOI:

https://doi.org/10.56211/sudo.v5i3.1998

Keywords:

Mikrogrid; Prediksi Daya; Random Forest; Rekayasa Fitur; Peramalan Jangka Pendek

Abstract

Ketersediaan listrik di wilayah terpencil Indonesia masih menghadapi tantangan akibat ketergantungan pada generator diesel dan sifat intermiten energi surya. Penelitian ini bertujuan mengembangkan model prediksi daya panel surya jangka pendek berbasis Random Forest multi-output dengan memanfaatkan data historis tiga jam untuk meramalkan keluaran daya pada horizon 30 menit ke depan. Data diperoleh dari sistem panel surya 50 Wp yang dilengkapi sensor arus, tegangan, dan suhu, direkam setiap 10 detik selama periode 1 bulan. Setelah pra-pemrosesan dan resampling ke resolusi satu menit, dilakukan rekayasa fitur meliputi komponen waktu siklik, lag features, statistik bergerak, serta diferensiasi. Dataset supervised dibentuk melalui pendekatan sliding window, di mana model memanfaatkan data historis selama 180 menit sebagai input untuk meramalkan keluaran daya pada horizon 30 menit ke depan. Kemudian model dilatih menggunakan MultiOutputRegressor dengan 100 pohon keputusan. Hasil evaluasi menunjukkan bahwa model mampu mencapai nilai R² hingga 0,94 pada sampel terbaik, dengan MAE dan RMSE yang masih berada pada rentang dapat diterima untuk aplikasi mikrogrid skala laboratorium. Prediksi pada horizon menit ke-5 hingga ke-15 cenderung lebih akurat dibandingkan horizon ke-30. Model ini menawarkan keseimbangan antara akurasi, interpretabilitas, dan efisiensi komputasi, sehingga berpotensi mendukung pengelolaan energi proaktif pada sistem mikrogrid berbasis surya di wilayah terpencil. Pengembangan lebih lanjut disarankan melalui penambahan variabel meteorologi eksternal dan pendekatan hibrida untuk meningkatkan kinerja pada kondisi cuaca dinamis.

Downloads

Download data is not yet available.

References

[1] T. Mosetlhe, A. Yusuff, T. Ayodele, dan A. Ogunjuyigbe, “Sustainable rural electrification through micro-grids in developing nations — A review of recent development,” Energy Reports, vol. 13, hlm. 1171–1177, Jun 2025, doi: 10.1016/J.EGYR.2024.11.040.

[2] M. A. Shehu dkk., “Control and energy management of standalone microgrids in remote areas: A review of recent advances, challenges, and opportunities for future research,” Engineering Science and Technology, an International Journal, vol. 75, hlm. 102288, Mar 2026, doi: 10.1016/J.JESTCH.2026.102288.

[3] W. M. Figueredo, O. R. Saavedra, P. B. Leite Neto, dan D. Q. Oliveira, “Design and operational challenges of renewable-powered isolated microgrids: a zero-carbon approach,” Scientific Reports 2025 15:1, vol. 15, no. 1, hlm. 45296-, Nov 2025, doi: 10.1038/s41598-025-28882-x.

[4] L. Vankudoth dan A. Q. H. Badar, “Development and analysis of scheduling strategies for utilizing shared energy storage system in networked microgrids,” J. Energy Storage, vol. 97, hlm. 112691, Sep 2024, doi: 10.1016/J.EST.2024.112691.

[5] M. M. Husein, E. J. Gago, B. Hasan, dan M. C. Pegalajar, “Towards energy efficiency: A comprehensive review of deep learning-based photovoltaic power forecasting strategies,” Heliyon, vol. 10, no. 13, hlm. e33419, Jul 2024, doi: 10.1016/J.HELIYON.2024.E33419.

[6] O. Khouili, M. Hanine, M. Louzazni, M. A. L. Flores, E. G. Villena, dan I. Ashraf, “Evaluating the impact of deep learning approaches on solar and photovoltaic power forecasting: A systematic review,” Energy Strategy Reviews, vol. 59, hlm. 101735, Mei 2025, doi: 10.1016/J.ESR.2025.101735.

[7] D. Hu, Y. Fan, dan W. Shao, “Optimal sizing and operation of microgrid considering renewable energy uncertainty based on scenario generation,” J. Energy Storage, vol. 109, hlm. 115174, Feb 2025, doi: 10.1016/J.EST.2024.115174.

[8] O. Motamedisedeh, S. Omrani, A. Karim, R. Drogemuller, dan G. Walker, “A comprehensive review of optimum integration of photovoltaic-based energy systems,” Renewable and Sustainable Energy Reviews, vol. 207, hlm. 114935, Jan 2025, doi: 10.1016/J.RSER.2024.114935.

[9] H. Z. Anonto dkk., “Intelligent energy management of microgrids using machine learning: Leveraging random forest models for solar and wind power,” Results in Engineering, vol. 27, hlm. 106539, Sep 2025, doi: 10.1016/J.RINENG.2025.106539.

[10] M. U. Tariq, S. M. Saqib, T. Mazhar, M. A. Khan, T. Shahzad, dan H. Hamam, “PV forecasting-driven decision-support for solar microgrids with battery storage,” Appl. Energy, vol. 412, hlm. 127685, Jun 2026, doi: 10.1016/J.APENERGY.2026.127685.

[11] M. Abdelsattar, M. A. Azim, A. AbdelMoety, dan A. Emad-Eldeen, “Comparative analysis of deep learning architectures in solar power prediction,” Scientific Reports 2025 15:1, vol. 15, no. 1, hlm. 31729-, Agu 2025, doi: 10.1038/s41598-025-14908-x.

[12] D. Sharma, C. Kumar, dan M. Kumar, “A review of machine learning and deep learning models, features, and applications for solar PV forecasting,” Journal of Electrical Systems and Information Technology 2026 13:1, vol. 13, no. 1, hlm. 74-, Jul 2026, doi: 10.1186/S43067-026-00382-6.

[13] C. Rodriguez-Aburto dkk., “Machine Learning for Photovoltaic Power Forecasting Integrated with Energy Storage Systems: A Scientometric Analysis, Systematic Review, and Meta-Analysis,” Energies 2025, Vol. 18, Page 6291, vol. 18, no. 23, hlm. 6291, Nov 2025, doi: 10.3390/EN18236291.

[14] F. Moazzen dan M. J. Hossain, “Multivariate Deep Learning Long Short-Term Memory-Based Forecasting for Microgrid Energy Management Systems,” Energies 2024, Vol. 17, Page 4360, vol. 17, no. 17, hlm. 4360, Agu 2024, doi: 10.3390/EN17174360.

[15] H. Xue, J. Ma, J. Zhang, P. Jin, J. Wu, dan F. Du, “Power Forecasting for Photovoltaic Microgrid Based on MultiScale CNN-LSTM Network Models,” Energies 2024, Vol. 17, Page 3877, vol. 17, no. 16, hlm. 3877, Agu 2024, doi: 10.3390/EN17163877.

[16] M. A. Azman, H. Jantan, U. F. M. Bahrin, dan E. A. Kadir, “Solar Power Production Forecasting Model Using Random Forest Algorithm,” Lecture Notes in Networks and Systems, vol. 1050 LNNS, hlm. 135–144, 2024, doi: 10.1007/978-3-031-64847-2_12/SAVE-RESEARCH.

[17] S. Jogunuri dkk., “Random forest machine learning algorithm based seasonal multi-step ahead short-term solar photovoltaic power output forecasting,” IET Renewable Power Generation, vol. 19, no. 1, hlm. e12921, Jan 2025, doi: 10.1049/RPG2.12921;PAGE:STRING:ARTICLE/CHAPTER.

[18] R. S. Kumar, P. S. Meera, V. Lavanya, dan S. Hemamalini, “Brown bear optimized random forest model for short term solar power forecasting,” Results in Engineering, vol. 25, hlm. 104583, Mar 2025, doi: 10.1016/J.RINENG.2025.104583.

[19] A. Masache, P. Mdlongwa, D. Maposa, dan C. Sigauke, “Short-term forecasting of solar irradiance using decision tree-based models and non-parametric quantile regression,” PLoS One, vol. 19, no. 12, hlm. e0312814, Des 2024, doi: 10.1371/JOURNAL.PONE.0312814.

[20] T. A. Nguyen dkk., “A Study on Novel Solar Power Forecasting using an XGB-LiGBM-RF Hybrid Model and the L-BFGS-B Optimization Algorithm,” Engineering, Technology & Applied Science Research, vol. 15, no. 4, hlm. 24516–24522, Agu 2025, doi: 10.48084/ETASR.11308.