A Study on the Impact of Categorical Alarm Data in Power Estimation and Anomaly Detection of Photovoltaic Inverters
##plugins.themes.bootstrap3.article.main##
##plugins.themes.bootstrap3.article.sidebar##
Abstract
This study investigates the impact of incorporating categorical inverter data into power-forecasting and anomaly-detection frameworks. Three forecasting models are evaluated on their ability to estimate power output on a large dataset coming from a fleet of multiple photovoltaic plants, over one hundred inverters and an approximate total of 33.2 installed MW. The forecasting models employed are Multi-Layer Perceptron, Long Short-Term Memory, and Extreme Gradient Boosting. Two encoding strategies for categorical alarm codes are compared: one-hot encoding and entity embeddings. Anomaly detection is performed by analysing residuals between predicted and measured power output. By systematically evaluating the integration of categorical inverter data into PV monitoring models,this work addresses an important gap in the literature and provides a foundation for future research exploring advanced methods for exploiting categorical operational data in photovoltaic systems.
How to Cite
##plugins.themes.bootstrap3.article.details##
photovoltaic, anomaly detection, predictive maintenance, scada, power forecasting, categorical alarm data, residual analysis, machine learning
Chang, Z., & Han, T. (2024). Prognostics and health management of photovoltaic systems based on deep learning: A state-of-the-art review and future perspectives. Renewable and Sustainable Energy Reviews, 205, 114861. doi: 10.1016/j.rser.2024.114861
Chen, T., & Guestrin, C. (2016). XGBoost: A scalable tree boosting system. In Proceedings of the 22nd ACM SIGKDD International Conference on Knowledge Discovery and Data Mining (pp. 785–794). doi: 10.1145/2939672.2939785
De Benedetti, M., Leonardi, F., Messina, F., Santoro, C., & Vasilakos, A. (2018). Anomaly detection and predictive maintenance for photovoltaic systems. Neurocomputing, 310, 59–68. doi: 10.1016/j.neucom.2018.05.017
Guo, C., & Berkhahn, F. (2016). Entity embeddings of categorical variables. arXiv. doi: 10.48550/arXiv.1604.06737
Gutschi, C., Furian, N., Suschnigg, J., Neubacher, D., & Voessner, S. (2019). Log-based predictive maintenance in discrete parts manufacturing. Procedia CIRP, 79, 528–533. doi: 10.1016/j.procir.2019.02.098
Hashemi, B., Taheri, S., Cretu, A.-M., & Pouresmaeil, E. (2021). Systematic photovoltaic system power losses calculation and modeling using computational intelligence techniques. Applied Energy, 284, 116396. doi: 10.1016/j.apenergy.2020.116396
Hochreiter, S., & Schmidhuber, J. (1997). Long short-term memory. Neural Computation, 9(8), 1735–1780. doi: 10.1162/neco.1997.9.8.1735
Ibrahim, M., Alsheikh, A., Awaysheh, F., & Alshehri, M. (2022). Machine learning schemes for anomaly detection in solar power plants. Energies, 15(3), 1082. doi: 10.3390/en15031082
International Energy Agency. (2025). World energy outlook 2025 (Tech. Rep.).
Kim, J., Kim, H., Kim, H., Lee, D., & Yoon, S. (2025). A comprehensive survey of deep learning for time series forecasting: Architectural diversity and open challenges. Artificial Intelligence Review, 58(7), 216. doi: 10.1007/s10462-025-11223-9
Liebermann, S., Um, J.-S., Hwang, Y., & Schlüter, S. (2021). Performance evaluation of neural network-based short-term solar irradiation forecasts. Energies, 14(11), 3030. doi: 10.3390/en14113030
Luo, M., Li, X., Zhang, D., Zhao, Y., & Lim, P. (2008). Categorical data analysis for equipment failure prediction. In 2008 34th Annual Conference of IEEE Industrial Electronics (pp. 1473–1478). IEEE. doi: 10.1109/IECON.2008.4758171
Marangis, D., Livera, A., Tziolis, G., Makrides, G., Kyprianou, A., & Georghiou, G. E. (2024). Trend-based predictive maintenance and fault detection analytics for photovoltaic power plants. Solar RRL, 8(24), 2400473. doi: 10.1002/solr.202400473
Onal, Y. (2022). Gaussian kernel-based SVR model for short-term photovoltaic MPP power prediction. Computer Systems Science and Engineering, 41(1), 141–156. doi: 10.32604/csse.2022.020367
Rumelhart, D. E., Hinton, G. E., & Williams, R. J. (1986). Learning representations by back-propagating errors. Nature, 323(6088), 533–536. doi: 10.1038/323533a0
Sheppard, S., Dickey, K. A., Koskey, S., Teasley, C., Perullo, C., Fregosi, D., & Li, W. (2024). Benchmarking a physics-based approach for anomaly detection at utility PV plants. In 2024 IEEE 52nd Photovoltaic Specialist Conference (PVSC) (pp. 856–858). IEEE. doi: 10.1109/PVSC57443.2024.10749158
Syamsuddin, A., Adhi, A. C., Kusumawardhani, A., Prahasto, T., & Widodo, A. (2024). Predictive maintenance based on anomaly detection in photovoltaic system using SCADA data and machine learning. Results in Engineering, 24, 103589. doi: 10.1016/j.rineng.2024.103589
Yao, S., Kang, Q., Zhou, M., Abusorrah, A., & Al-Turki, Y. (2021). Intelligent and data-driven fault detection of photovoltaic plants. Processes, 9(10), 1711. doi: 10.3390/pr9101711
This work is licensed under a Creative Commons Attribution 3.0 Unported License.
The Prognostic and Health Management Society advocates open-access to scientific data and uses a Creative Commons license for publishing and distributing any papers. A Creative Commons license does not relinquish the author’s copyright; rather it allows them to share some of their rights with any member of the public under certain conditions whilst enjoying full legal protection. By submitting an article to the International Conference of the Prognostics and Health Management Society, the authors agree to be bound by the associated terms and conditions including the following:
As the author, you retain the copyright to your Work. By submitting your Work, you are granting anybody the right to copy, distribute and transmit your Work and to adapt your Work with proper attribution under the terms of the Creative Commons Attribution 3.0 United States license. You assign rights to the Prognostics and Health Management Society to publish and disseminate your Work through electronic and print media if it is accepted for publication. A license note citing the Creative Commons Attribution 3.0 United States License as shown below needs to be placed in the footnote on the first page of the article.
First Author et al. This is an open-access article distributed under the terms of the Creative Commons Attribution 3.0 United States License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.