Light & Engineering 34 (1) 2026

Volume 34
Date of publication 02/19/2026
Pages 95–103

Deep Learning-Based Power Forecasting for Smart Street Lighting: A Comparative Study of LSTM and CNN-LSTM L&E, Vol.34, No.1, 2026

Articles authors:

Mouaadh Yaichi, Zakaria Benarbia, Mhamed Rebhi, Bousmaha Bouchiba

Mouaadh Yaichi, M. Sc. In 2016, he graduated from Tahri Mohamed University-Bechar, Algeria, with a B. Sc. in electronic. In 2018, he received M. Sc. in electronic of embedded system from the same university. He is currently Ph. D. student in the field of smart lighting

Zakaria Benarbia, Ph. D. student in Machine Learning and Deep Learning for Computer Vision. He has got his M. Sc. in Electronic and Embedded Systems Engineering from Tahri Mohamed University. His research interests include deep learning, computer vision, and edge AI, with a particular focus on multi-object tracking using vision-based methods and deploying optimized AI models on embedded devices. By combining theoretical expertise with practical implementation, he aims to develop efficient, explainable, and robust AI systems for real-world applications

Mhamed Rebhi, Assistant Professor, Ph. D. of energetic physics from Tahri Mohamed University of Bechar, Algeria in 2017. He received M. Sc. of photovoltaic microelectronics in 2007, and graduated in micro- electronics engineering in 1993 from university of SDB. Currently, he is an Assistant Professor at Tahri Mohamed University of Bechar and member in the Laboratory of Smart Grids & Renewable Energies. His area of research interest includes modelling, sizing and optimization of photovoltaic systems, energy management, conventional & intelligent MPPT algorithms, embedded electronic, and microgrid systems

Bousmaha Bouchiba, Ph. D., Prof. He received the Electrical Engineering diploma from Tahri Mohamed University-Bechar, Algeria in 1999, M. Sc. from the University Alexandria Egypt in 2006, and the Ph. D. from Djillali Liabes University of Sidi Bel Abbes in 2011. Currently, he is the Professor of electrical engineering at Tahri Mohamed University-Bechar, Algeria

Abstract:

Southern Algeria is distinguished by severe climatic conditions, notably persistent strong winds and extreme temperatures. These environmental factors frequently cause malfunctions in smart street lighting systems, leading to operational downtime that adversely affects the daily lives of citizens. During pilot testing of a smart street lighting system on a university campus, two critical challenges were identified: strong winds and heavy rainfall, both of which forced the system to suspend operation due to reduced performance under such conditions. To mitigate these issues, two deep learning models – Long Short-Term Memory (LSTM) and Convolutional Neural Network-LSTM (CNN-LSTM) – were trained and evaluated using a univariate dataset comprising only historical power output. The primary objective was to design a control strategy capable of dynamically regulating power levels to improve the system’s resilience to environmental variability. The findings indicate that both models delivered robust predictive performance; however, the LSTM model achieved superior accuracy compared to the CNN-LSTM, highlighting its effectiveness in optimizing smart street lighting systems under extreme climatic conditions while simultaneously promoting energy efficiency.

References:

1. M’hamed, R., Boussemaha, B., Mouaadh, Y., Youcef, H., Mohammed, D. Low cost of smart unilateral street lighting system with photovoltaic power plant using particle swarm optimization // Studies in Engineering and Exact Sciences, 2024, Vol. 5, # 2, p. e11658.
2. Agramelal, F., Sadik, M., Moubarak, Y., Abouzahir, S. Smart Street Light Control: A Review on Methods, Innovations, and Extended Applications // Energies, 2023, Vol. 16, # 21, pp. 1–42.
3. Jackett, M., Frith, W. Quantifying the impact of road lighting on road safety – A New Zealand Study // IATSS Research., 2013, Vol. 36, # 2, pp. 139–145.
4. Viswanathan, S., Momand, S., Fruten, M., Alcantar, A. A model for the assessment of energy-efficient smart street lighting – a case study // Energy Efficiency, 2021, Vol. 14, # 6.
5. Pizzuti, S., Annunziato, M., Moretti, F. Smart Street Lighting Management // Energy Efficiency, 2013. V6, # 3, pp. 607–616.
6. Beccali, M., et al. A multifunctional public lighting infrastructure, design and experimental test // Journal of Sustainable Development of Energy, Water, and Environment Systems, 2017, Vol. 5, # 4, pp. 608–625.
7. Rojek, I., Mikołajewski, D., Galas, K., Piszcz, A. Advanced Deep Learning Algorithms for Energy Optimization of Smart Cities // Energies, 2025, Vol. 18, # 2.
8. Raza, M.Q., Khosravi, A. A review on artificial intelligence-based load demand forecasting techniques for smart grid and buildings // Renewable and Sustainable Energy Reviews, 2015, Vol. 50, pp. 1352–1372.
9. Alam, S., et al. Intelligent Streetlight Control System Using Machine Learning Algorithms for Enhanced Energy Optimization in Smart Cities // Journal of Ecohumanism, 2025, Vol. 4, # 4.
10. Durand, D., Aguilar, J., R-Moreno, M.D. An analysis of the energy consumption forecasting problem in smart buildings using LSTM // Sustainability, 2022, Vol. 14, # 20.
11. Jouahri, M.A., Boulghasoul, Z., Tajer, A. Smart control of electrical power in an LED lighting network taking into account road flow and meteorological conditions // Electrical Engineering, 2024. Vol. 106, # 5, pp. 5655–5675.
12. Gopalakrishnan, B., Dharshini, M., Devi, S.R., Gayathri, K. Prediction of street lights in metro cities using time series analysis in machine learning algorithm // International Journal of Aquatic Science, 2021. Vol. 12, # 3, pp. 1336–1345.
13. Amiri, A.F., Chouder, A., Oudira, H., Silvestre, S., Kichou, S. Improving photovoltaic power prediction: insights through computational modelling and feature selection // Energies, 2024, Vol. 17, # 13.
14. Nadweh, S., Mohammed, N., Konstantinou, C., Ahmed, S. Operational performance assessment of PVpowered street lighting: a comparative study of different machine learning prediction models // IEEE Access, 2025, Vol. 13.
15. Yang, G., Luo, Q., Wu, J. Automatic control method for street lights in unideal lighting environments based on deep learning // Light & Engineering., 2023, Vol. 31, # 6, pp. 100–108.
16. Bian, J., Yang, J.J. Smart Street lighting powered by renewable energy: a multi-criteria, data-driven decision framework // Sustainability, 2025, Vol. 17, # 13.
17. Sujana, J.A.J., Raj, R.V., Priya, V.K.R. Fog intelligence for energy efficient management in smart street lamps // Computing, 2024, Vol. 106, # 12, pp. 4057–4082.
18. Mouaadh, Y., Bousmaha, B., Mhamed, R. Intelligent control and reduce energy consumption of smart street lighting system // International Journal of Power Electronics and Drive Systems (IJPEDS), 2022, Vol. 13, # 4, pp. 1966–1974.
19. Arpit, D., Kanuparthi, B., Kerg, G., Ke, N.R., Mitliagkas, I., Bengio, Y. h-detach: Modifying the LSTM gradient towards better optimization // arXiv Prepr., 2018. arXiv:1810.03023.
20. Elmaz, F., Eyckerman, R., Casteels, W., Latré, S., Hellinckx, P. CNN-LSTM architecture for predictive indoor temperature modelling // Building and Environment, 2021, Vol. 206, p.108327.
21. Kim, T.Y., Cho, S.B. Predicting residential energy consumption using CNN-LSTM neural networks // Energy, 2019, Vol. 182, pp. 72–81.
22. Lee, S.W., Kim, H.Y. Stock market forecasting with super-high dimensional time-series data using ConvLSTM, trend sampling, and specialized data augmentation // Expert Systems with Applications, 2020, Vol. 161, p.113704.
23. Tukymbekov, D., Saymbetov, A., Nurgaliyev, M., Kuttybay, N., Dosymbetova, G., Svanbayev, Y. Intelligent autonomous street lighting system based on weather forecast using LSTM // Energy, 2021, Vol. 231, p.120902.
24. Islam, M.Z., Islam, M.M., Asraf, A. A combined deep CNN-LSTM network for the detection of novel coronavirus (COVID‑19) using X-ray images // Informatics in Medicine Unlocked, 2020, Vol. 20, p.100412.
25. Wang, J., Yu, L.C., Lai, K.R., Zhang, X. Dimensional sentiment analysis using a regional CNN-LSTM model // Proc. 54th Annu. Meet. Assoc. Comput. Linguist. (ACL), 2016, pp. 225–230.
26. Yaichi, M., Bouchiba, B., Rebhi, M. IoT-based smart street lighting system and enhanced energy saving using ESP-NOW protocol // Light & Engineering, 2024, Vol. 32, # 6, pp. 76–84.
27. Agga, A., Abbou, A., Labbadi, M., El Houm, Y. Short-term self-consumption PV plant power production forecasts based on hybrid CNN-LSTM, ConvLSTM models // Renewable Energy, 2021, Vol. 177, pp. 101–112.
28. Tsalikidis, N., Mystakidis, A., Tjortjis, C., Koukaras, P., Ioannidis, D. Energy load forecasting: one-step ahead hybrid model utilizing ensembling // Computing, 2024. Vol. 106, # 1, pp. 1–33.

Keywords

DOI: https://doi.org/10.33383/2025-065

Article in RUS:
Svetotekhnika # 1, 2026

Buy