Authors : Shubhangi Darade; Suresh R. Jajoo

Volume/Issue : Volume 11 - 2026, Issue 2 - February

Google Scholar : https://tinyurl.com/mup6dm28

Scribd : https://tinyurl.com/4s95y9ks

DOI : https://doi.org/10.38124/ijisrt/26feb1293

Note : A published paper may take 4-5 working days from the publication date to appear in PlumX Metrics, Semantic Scholar, and ResearchGate.

Abstract : The growing adoption of photovoltaic (PV) systems into contemporary power networks requires proper forecasting of the generation of solar power and effective fault identification strategies to make sure that the system can operate reliably and produce energy when it is required. The nature of solar energy production is intermittent because it is highly conditioned by the characteristics of the environment including sun radiation, ambient temperature, and temperature of the modules. The paper includes a detailed machine learning-powered prediction of solar power generation and anomaly detection models based on real-world working data of a photovoltaic power station. Parameters of power generation such as DC power, AC power, daily yield, and total yield and weather sensor parameters such as irradiation, ambient temperature, and module temperature constitute the dataset. Three regression models (Linear Regression, Decision Tree Regressor, and Random Forest Regressor) were adopted in predicting AC power output after a preprocessing and feature extraction step. R2 score, Mean Absolute Error (MAE) and Root Mean Square Error (RMSE) were used to assess the model performance. The experimental findings show that the Random Forest Regressor performed well than the other models, and the R2 score is 92.7, which shows that it has a high predictive ability and resilience to nonlinear deviations. Furthermore, there was a risk of detecting anomalies by daily patterns of power generation, and this approach allowed identifying unusual fluctuations with the main causes of which are environmental disturbances like cloud cover and high temperatures. The analysis of the inverter efficiency showed that the overall efficiency was around 93 percent with slight decreases in the conditions of extreme temperature. The results prove the usefulness of the machine learning methods in improving the accuracy of solar power prognostics, identifying the operational aberration, and assisting the predictive maintenance approaches in order to attain better photovoltaic systems reliability and grid stability.

Keywords : Solar Power Forecasting, Photovoltaic Systems, Machine Learning, Random Forest Regression, Anomaly Detection, Inverter Efficiency.

References :

1. M. Mamta and A. Kumar, "Shining a Light on Solar Power Forecasting: Machine Learning Techniques for Unprecedented Accuracy," Proc. of the Conference Name, Jan. 2024. [Online]. Available: https://www.researchgate.net/publication/379350977_Solar_Power_Forecasting_using_Machine_Learning_Approaches_A_Review
2. D. Singh, "Machine Learning Predictions for Fault Detections in Solar Photovoltaic System: A Bibliographic Outlook," Journal Name, vol. 15, no. 3, pp. 123-135, Mar. 2025. [Online]. Available: https://www.researchgate.net/publication/389780489_Machine_learning_predictions_for_fault_detections_in_solar_photovoltaic_system_A_bibliographic_outlook/download
3. A. Sharma, P. Gopi, and K. Sudhakar, "Weather Impact on Solar Farm Performance: A Comparative Analysis of Machine Learning Techniques," Sustainability, vol. 15, no. 1, pp. 439-452, Jan. 2023. [Online]. Available: https://doi.org/10.3390/su15010439
4. Y. Hajji, M.-F. Harkat, and A. Kouadri, "Multivariate Feature Extraction Based Supervised Machine Learning for Fault Detection and Diagnosis in Photovoltaic Systems," European Journal of Control, vol. 59, pp. 313-321, Mar. 2021. [Online]. Available: https://doi.org/10.1016/j.ejcon.2020.03.004
5. F. Harrou, A. Saidi, and Y. Sun, "Monitoring of Photovoltaic Systems Using Improved Kernel-Based Learning Schemes," IEEE Journal of Photovoltaics, vol. 11, no. 3, pp. 806-818, May 2021. [Online]. Available: https://doi.org/10.1109/jphotov.2021.3057169
6. C. He, J. Liu, and F. Xu, "Improving Solar Radiation Estimation in China Based on Regional Optimal Combination of Meteorological Factors with Machine Learning Methods," Energy Conversion and Management, vol. 220, p. 113111, Aug. 2020. [Online]. Available: https://doi.org/10.1016/j.enconman.2020.113111
7. S. Hussain and A. AlAlili, "A Hybrid Solar Radiation Modeling Approach Using Wavelet Multiresolution Analysis and Artificial Neural Networks," Applied Energy, vol. 208, pp. 540-550, Dec. 2017. [Online]. Available: https://doi.org/10.1016/j.apenergy.2017.09.100
8. I. A. Ibrahim and T. Khatib, "A Novel Hybrid Model for Hourly Global Solar Radiation Prediction Using Random Forests Technique and Firefly Algorithm," Energy Conversion and Management, vol. 138, pp. 413-425, Mar. 2017. [Online]. Available: https://doi.org/10.1016/j.enconman.2017.02.006
9. R. Ismaen, M. Kucukvar, and T. Y. El Mekkawy, "Optimization and Enviro-Economic Assessment of Solar-Cooling Systems Towards Sustainable Development: A Case Study of Qatar," Journal of Cleaner Production, vol. 419, p. 138253, Jan. 2023. [Online]. Available: https://doi.org/10.1016/j.jclepro.2023.138253
10. E. İzgi, A. Öztopal, and B. Yerli, "Short–Mid-Term Solar Power Prediction by Using Artificial Neural Networks," Solar Energy, vol. 86, no. 2, pp. 725-733, Feb. 2012. [Online]. Available: https://doi.org/10.1016/j.solener.2011.11.013
11. A. Jäger-Waldau, "Snapshot of Photovoltaics—February 2022," EPJ Photovoltaics, vol. 13, p. 9, Feb. 2022. [Online]. Available: https://doi.org/10.1051/epjpv/2022010
12. S. Jain and P. K. Jain, "The Rise of Renewable Energy Implementation in South Africa," Energy Procedia, vol. 143, pp. 721-726, Dec. 2017. [Online]. Available: https://doi.org/10.1016/j.egypro.2017.12.752
13. H. S. Jang, K. Y. Bae, and H.-S. Park, "Solar Power Prediction Based on Satellite Images and Support Vector Machine," IEEE Transactions on Sustainable Energy, vol. 7, no. 3, pp. 1255-1263, Jul. 2016. [Online]. Available: [https://doi.org/10.1109/tste.2016.2535466](https://doi.org/
14. M. Aghaei, "Fault Detection for Photovoltaic Systems Using Machine Learning Techniques," presented at the Training School III: Simulation Tools and Models for the Analysis of PV System Performance, Brasov, Romania, Jul. 2021. [Online]. Available: https://www.pearlpv-cost.eu/wp-content/uploads/2020/05/Aghei-Fault-Detection-for-Photovoltaic-Systems-using-Machine-Learning-Techniques_07072021.pdf
15. L. Bommes et al., "Anomaly Detection in IR Images of PV Modules Using Supervised Contrastive Learning," arXiv preprint arXiv:2112.02922, Dec. 2021. [Online]. Available: https://arxiv.org/abs/2112.02922
16. N. Salazar-Pena, A. Tabares, and A. Gonzalez-Mancera, "AI-Powered Dynamic Fault Detection and Performance Assessment in Photovoltaic Systems," arXiv preprint arXiv:2409.00052, Aug. 2024. [Online]. Available: https://arxiv.org/abs/2409.00052
17. A. Prabhakar, "Solar Power Forecasting Using Machine Learning and Deep Learning," Medium, Aug. 2021. [Online]. Available: https://medium.com/@akashprabhakar427/solar-power-forecasting-using-machine-learning-and-deep-learning-61d6292693de
18. S. Hussain and A. AlAlili, "A Hybrid Solar Radiation Modeling Approach Using Wavelet Multiresolution Analysis and Artificial Neural Networks," Applied Energy, vol. 208, pp. 540-550, Dec. 2017. [Online]. Available: https://doi.org/10.1016/j.apenergy.2017.09.100