Artificial Intelligence and Machine Learning for Fault Detection and Energy Forecasting in Photovoltaic Systems: A Comprehensive Review

  • Gajanan Shravan Datar Pimpri Chinchwad College of Engineering affiliated with Savitribai Phule Pune University
  • Chandrakishor L. Ladekar Pimpri Chinchwad College of Engineering affiliated with Savitribai Phule Pune University
Keywords: Artificial Intelligence, Digital Twin, Fault Diagnosis, Machine Learning, Photovoltaic Systems, Renewable Energy.

Abstract

Worldwide installation of PV systems has increased trash demand for efficient monitoring and energy forecasting. Efficiency, safety, and financial risk management are the basic cornerstones considered while monitoring PV systems. Classical approaches for fault diagnosis and power prediction of PVs have become obsolete due to their limitations in handling nonlinearities under uncertainties and scalability under varying operational conditions. With the evolution of artificial intelligence and machine learning, intelligent datadriven frameworks can be developed for real-time fault diagnosis, performance evaluation, and predictive maintenance. This study intends to present a critical review of the latest AI and ML techniques in PV system monitoring and forecasting, addressing issues relating to their aptitude in the identification of the most common faults, such as hotspots, partial shading, soiling, and inverter failures, together with improving short- and long-term energy prediction. Deep learning and hybrid AI models, which consider accuracy, sensitivity, and robustness across heterogeneous datasets, are far superior to traditional methods. Also, when integrated with IoT, edge computing, and digital twin technologies, they build on scalability, adaptability, and decision-making capabilities in real time. The review also highlighted concerning issues of data scarcity, generalizability across different climates, explainability, and cybersecurity. Finally, future directions are outlined to create standard datasets and benchmarking practices and construct explainable hybrid models with a trustworthy and transparent foundation, further leading to the wide adoption of AI in PV systems.

References

[1] A. Thakfan and Y. Bin Salamah, “Artificial-intelligence-based detection of defects and faults in photovoltaic systems: A survey,” Energies, vol. 17, no. 19, p. 4807, 2024.
[2] A. Mellit and S. Kalogirou, “Assessment of machine learning and ensemble methods for fault diagnosis of photovoltaic systems,” Renewable Energy, vol. 184, pp. 1074–1090, 2022.
[3] G. Karthikeyan and A. Jagadeeshwaran, “Enhancing solar energy generation: A comprehensive machine learning-based PV prediction and fault analysis system for real-time tracking and forecasting,” Electric Power Components and Systems, vol. 52, no. 9, pp. 1497– 1512, 2024.
[4] A. Abubakar, M. M. Jibril, C. F. Almeida, M. Gemignani, M. N. Yahya, and S. I. Abba, “A novel hybrid optimization approach for fault detection in photovoltaic arrays and inverters using AI and statistical learning techniques: A focus on sustainable environment,” Processes, vol. 11, no. 9, p. 2549, 2023.
[5] G. M. El-Banby, N. M. Moawad, B. A. Abouzalm, W. F. Abouzaid, and E. A. Ramadan, “Photovoltaic system fault detection techniques: A review,” Neural Computing and Applications, vol. 35, no. 35, pp. 24829–24842, 2023.
[6] A. Mohammad and F. Mahjabeen, “Revolutionizing solar energy: The impact of artificial intelligence on photovoltaic systems,” International Journal of Multidisciplinary Sciences and Arts, vol. 2, no. 3, p. 591856, 2023.
[7] A. A. Al-Katheri, E. A. Al-Ammar, M. A. Alotaibi, W. Ko, S. Park, and H. J. Choi, “Application of artificial intelligence in PV fault detection,” Sustainability, vol. 14, no. 21, p. 13815, 2022.
[8] S. Voutsinas, D. Karolidis, I. Voyiatzis, and M. Samarakou, “Development of a machine-learning-based method for early fault detection in photovoltaic systems,” Journal of Engineering and Applied Science, vol. 70, no. 1, p. 27, 2023.
[9] L. D. Jathar, K. Nikam, U. V. Awasarmol, R. Gurav, J. D. Patil, K. Shahapurkar, et al., “A comprehensive analysis of the emerging modern trends in research on photovoltaic systems and desalination in the era of artificial intelligence and machine learning,” Heliyon, vol. 10, no. 3, 2024.
[10] A. Mellit, M. Benghanem, S. Kalogirou, and A. M. Pavan, “An embedded system for remote monitoring and fault diagnosis of photovoltaic arrays using machine learning and the internet of things,” Renewable Energy, vol. 208, pp. 399–408, 2023.
[11] S. Deitsch et al., “Automatic classification of defective photovoltaic module cells in electroluminescence images,” Solar Energy, vol. 185, pp. 455–468, 2019.
[12] J. Zhou and Y. Zhang, “Short-term photovoltaic power forecasting based on LSTM with attention,” IEEE Access, vol. 7, pp. 751–761, 2019.
[13] P. Li et al., “A hybrid deep learning model for short-term PV power forecasting,” Applied Energy, vol. 260, p. 114–121, 2020.
[14] D. Niu et al., “Short-term PV power forecasting based on RF feature selection and CEEMD,” Applied Soft Computing, vol. 92, p. 106293, 2020.
[15] H. Zang et al., “Day-ahead PV power forecasting via deep CNNs and meta-learning,” International Journal of Electrical Power & Energy Systems, vol. 118, p. 105790, 2020.
[16] P. S. Paradell et al., “Photovoltaics plant fault detection using deep learning techniques,” Remote Sensing, 2022.
[17] Y. Li et al., “Short-term PV power forecasting based on historical data with deep learning,” Sensors, 2022.
[18] R. Bessa et al., “Short-term PV forecasting using meta-learning and dynamic blending,” Renewable Energy, 2023.
[19] B. Gu et al., “Forecasting and uncertainty analysis of day-ahead PV power based on WT-CNN-BiLSTM-AM-GMM,” Sustainability, 2023.
[20] X. Zhang et al., “Digital twin approach for fault diagnosis in PV plant DC–DC converters,” Sensors, 2025.
[21] Y. Y. Hong and R. A. Pula, “Diagnosis of PV faults using digital twin and PSO-optimized shifted-window Transformer,” Applied Soft Computing, 2023/2024.
[22] A. Abdelrahman et al., “Faults detection and diagnosis of PV systems based on ML,” Energy Conversion and Management, 2024.
[23] M. Alrashidi et al., “Short-term PV power forecasting based on temporal CNN–GRU (TCN-ECANet-GRU),” PLOS ONE, 2024.
[24] A. Abdelkader et al., “Diagnosing faults in a PV system using Extra Trees,” AIMS Energy, 2024.
[25] Y. Tao et al., “PV array fault diagnosis and localization via modulated photocurrent and ML,” Sensors, 2025.
[26] Y. Gaaloul et al., “Faults detection and diagnosis of a large-scale PV system via RF and k-NN predictive models,” Energies, 2025.
[27] M. Rahman et al., “Leveraging U-Net and ASPP for effective fault detection in PV systems,” Scientific Reports, 2025.
[28] S. Deitsch et al., “Automatic classification of defective PV module cells in EL images,” Solar Energy, 2019.
[29] Y. Jin et al., “CNN-based automatic detection of PV cell defects in EL images,” Energy, 2019.
[30] W. Tang, Q. Yang, K. Xiong, and W. Yan, “Deep learning-based automatic defect identification of PV modules using EL images,” Solar Energy, 2020.
[31] P. Li et al., “A hybrid deep learning model (WPD-LSTM) for shortterm PV power forecasting,” Applied Energy, 2020.
[32] R. Sharma et al., “PV cell defect classification using CNNs,” IET Renewable Power Generation, 2020.
[33] P. S. Paradell et al., “Photovoltaics plant fault detection using deep learning (UAV thermal),” Remote Sensing, 2022.
[34] Y. Li et al., “Short-term PV power forecasting with deep learning,” Sensors, 2022.
[35] Z. Cao et al., “Lightweight hotspot fault detection model of PV panels,” Sensors, 2022.
[36] R. Bessa et al., “Short-term PV forecasting using meta-learning and dynamic blending,” Renewable Energy, 2023.
[37] X. Wang et al., “Short-term PV power forecasting via time-series expansion and IVMD hybridization,” Applied Soft Computing, 2023.
[38] H. Liu et al., “Defect detection of PV modules based on improved UNet,” Scientific Reports, 2024.
[39] A. Sousa et al., “Short-term PV energy forecasting with LSTM on public EDP data,” Energies, 2024.
[40] S. Zhang et al., “Short-term PV probabilistic forecasting via decomposition and Vine Copula,” Sustainability, 2024.
[41] M. Rahman et al., “Leveraging U-Net and ASPP for effective fault detection in PV systems,” Scientific Reports, 2025.
[42] J. Chen et al., “Transformer-based deep neural networks for short-term solar power forecasting,” Engineering Applications of Artificial Intelligence, 2025.
[43] A. Chouder, S. Silvestre, N. S. Bailek, and L. El Ouariachi, “Development of an algorithm for fault detection in grid-connected PV systems,” Solar Energy, vol. 159, pp. 650–659, Apr. 2018, doi: 10.1016/j.solener.2017.11.052.
[44] G. Massi Pavan, A. Mellit, and D. F. Corona, “A comparative study of different techniques for detection of defects in PV panels using infrared images,” Renewable Energy, vol. 123, pp. 362–373, Aug. 2018, doi: 10.1016/j.renene.2018.02.061.
[45] P. Sharma and M. Chandel, “A review and comparative analysis of solar photovoltaic array fault detection techniques,” Solar Energy, vol. 180, pp. 210–228, Mar. 2019, doi: 10.1016/j.solener.2019.01.039.
[46] F. Khan, S. Singh, and B. C. Pal, “An integrated approach for detection and localization of multiple faults in photovoltaic arrays,” IEEE Transactions on Sustainable Energy, vol. 11, no. 1, pp. 346–356, Jan.
2020, doi: 10.1109/TSTE.2019.2906175.
[47] A. Arab, A. Rahmann, and A. Yazdani, “Detection of faults in photovoltaic systems using machine learning and statistical techniques,” IEEE Transactions on Power Delivery, vol. 35, no. 4, pp. 1949–1958, Aug. 2020, doi: 10.1109/TPWRD.2019.2937438.
[48] L. Wang, J. Zhang, and M. Yang, “A CNN-based approach for detecting and classifying photovoltaic faults using infrared images,” Energies, vol. 13, no. 3, p. 620, Feb. 2020, doi: 10.3390/en13030620.
[49] H. Mekki, A. Mellit, and M. Benghanem, “Artificial neural networkbased modelling and fault detection of photovoltaic modules,” Renewable Energy, vol. 145, pp. 297–307, Jan. 2020, doi: 10.1016/j.renene.2019.06.054.
[50] S. Muniraj, D. Srinivasan, and J. Lu, “Deep learning-based PV fault detection with long short-term memory networks,” IEEE Access, vol.
8, pp. 120871–120884, Jul. 2020, doi:
10.1109/ACCESS.2020.3006198.
[51] Z. Zhao, L. Xu, and T. Yu, “Fault detection in photovoltaic systems using support vector machines and monitoring data,” Renewable
Energy, vol. 164, pp. 444–456, Feb. 2021, doi:
10.1016/j.renene.2020.09.092.
[52] F. Riganti Fulginei and A. Salvini, “A deep learning approach for the fault diagnosis in photovoltaic plants,” Applied Energy, vol. 275, p.
115390, Oct. 2020, doi: 10.1016/j.apenergy.2020.115390.
[53] H. Peng, Y. Sun, and X. Zhang, “Hybrid CNN–LSTM model for PV fault diagnosis using electrical data,” Energy Reports, vol. 7, pp. 2514– 2525, Nov. 2021, doi: 10.1016/j.egyr.2021.04.054.
[54] A. P. Marquez, R. M. Ferreira, and E. A. Silva, “Application of random forest for PV fault detection using SCADA data,” Electric Power
Systems Research, vol. 192, p. 106980, May 2021, doi:
10.1016/j.epsr.2020.106980.
[55] S. L. Xie, W. Gao, and H. Yang, “Probabilistic solar forecasting with LSTM networks and Gaussian processes,” Applied Energy, vol. 304, p. 117807, Jan. 2022, doi: 10.1016/j.apenergy.2021.117807.
[56] M. Esposito, D. Di Lorenzo, and L. Foggia, “An ensemble learning method for PV fault detection and performance prediction,” IEEE Transactions on Industrial Electronics, vol. 70, no. 2, pp. 1561–1571, Feb. 2023, doi: 10.1109/TIE.2022.3146704.
[57] R. AlShahrani, A. R. Al-Ali, and A. Abu-Siada, “AI-enabled frameworks for integrated PV fault detection and forecasting,” Renewable and Sustainable Energy Reviews, vol. 175, p. 113174, Mar. 2025, doi: 10.1016/j.rser.2024.113174.
Published
2025-12-10
How to Cite
Datar, G., & Ladekar, C. (2025). Artificial Intelligence and Machine Learning for Fault Detection and Energy Forecasting in Photovoltaic Systems: A Comprehensive Review. Asian Journal For Convergence In Technology (AJCT) ISSN -2350-1146, 11(1), 19-24. Retrieved from http://www.asianssr.org/index.php/ajct/article/view/1425

Most read articles by the same author(s)

Obs.: This plugin requires at least one statistics/report plugin to be enabled. If your statistics plugins provide more than one metric then please also select a main metric on the admin's site settings page and/or on the journal manager's settings pages.