Ultra-short-term power prediction of photovoltaic power systems based on multi-climatic environments
DOI:
https://doi.org/10.70917/jcc-2026-002Keywords:
photovoltaic power generation; ultra-short-term power prediction; deep learning; MHSA-LSTM; multi-climatic environmentAbstract
Ultra-short-term power prediction of photovoltaic (PV) power generation system is an important basis for grid scheduling and energy management. In order to adapt to the complex and changing climatic conditions, this paper proposes an MHSA-LSTM ultra-short-term power prediction model for PV power generation system that integrally considers all kinds of climatic factors. In this study, the raw data of a PV power plant is processed by multiple interpolation, and four key climate environment variables, namely, temperature, irradiance, relative humidity and atmospheric pressure, which affect the power generation, are extracted from them. The time series data containing the influencing variables are used as samples, and the MHSA module is utilized to filter the importance of the data at different historical moments, and the strong and weak weights of each environmental variable are adaptively assigned to the LSTM network to obtain the prediction results. The results show that climatic factors such as solar irradiance, temperature, relative humidity and atmospheric pressure can affect the PV power, with irradiance having the most significant effect.The MHSA-LSTM model is oriented to the climatic conditions of different seasons and weather, and the predicted ultrashort-term power is closer to the actual power value. Taking spring conditions as an example, the MAPE\RMSE of the MHSA-LSTM coupled model is reduced by 10.33%\2.771kW, 23.12%\6.296kW, 5.34%\1.234kW, compared to BP-LSTM, LSTM and LSTNet, respectively.The fluctuation of the ultra-short-term power prediction model based on deep learning is basically the same as that of the actual fluctuation. It is more adapted to the subsequent grid power scheduling and operation requirements.
References
Abdullah, B. U. D., Khanday, S. A., Islam, N. U., Lata, S., Fatima, H., & Nengroo, S. H. (2024). Comparative analysis using multiple regression models for forecasting photovoltaic power generation. Energies, 17(7), 1564.
Banik, R., & Biswas, A. (2023). Improving solar PV prediction performance with RF-CatBoost ensemble: a robust and complementary approach. Renewable Energy Focus, 46, 207-221.
Huang, C., & Yang, M. (2023). Memory long and short term time series network for ultra-short-term photovoltaic power forecasting. Energy, 279, 127961.
Jagadeesh, V., Venkata Subbaiah, K., & Varanasi, J. (2020). Forecasting the probability of solar power output using logistic regression algorithm. Journal of Statistics and Management Systems, 23(1), 1-16.
Jo, H. H., Kim, J., & Kim, S. (2024). Enhancing the power generation performance of photovoltaic system: Impact of environmental and system factors. Applied Thermal Engineering, 240, 122221.
Khan, Z. A., Hussain, T., & Baik, S. W. (2023). Dual stream network with attention mechanism for photovoltaic power forecasting. Applied Energy, 338, 120916.
Li, L., Lin, J., Wu, N., Xie, S., Meng, C., Zheng, Y., ... & Zhao, Y. (2022). Review and outlook on the international renewable energy development. Energy and Built Environment, 3(2), 139-157.
Liu, M., Wang, X., & Zhong, Z. (2025). Ultra-short-term photovoltaic power prediction based on BiLSTM with wavelet decomposition and dual attention mechanism. Electronics, 14(2), 306.
Mahesh, P. V., Meyyappan, S., & Alla, R. R. (2022). Maximum power point tracking using decision-tree machine-learning algorithm for photovoltaic systems. Clean Energy, 6(5), 762-775.
Niu, D., Wang, K., Sun, L., Wu, J., & Xu, X. (2020). Short-term photovoltaic power generation forecasting based on random forest feature selection and CEEMD: A case study. Applied soft computing, 93, 106389.
Rinesh, S., Deepa, S., Nandan, R. T., Sachin, R. S., Thamil, S. V., Akash, R., ... & Kumar, A. S. (2024). Prediction and classification of solar photovoltaic power generation using extreme gradient boosting regression model. International Journal of Low-Carbon Technologies, 19, 2420-2430.
Saxena, N., Kumar, R., Rao, Y. K., Mondloe, D. S., Dhapekar, N. K., Sharma, A., & Yadav, A. S. (2024). Hybrid KNN-SVM machine learning approach for solar power forecasting. Environmental Challenges, 14, 100838.
Souza, G., Santos, R., & Saraiva, E. (2022). A Log-logistic predictor for power generation in photovoltaic systems. Energies, 15(16), 5973.
Su, Z., Gu, S., Wang, J., & Lund, P. D. (2025). Improving ultra-short-term photovoltaic power forecasting using advanced deep-learning approach. Measurement, 239, 115405.
Tang, H., Kang, F., Li, X., & Sun, Y. (2025). Short-term photovoltaic power prediction model based on feature construction and improved transformer. Energy, 320, 135213.
Tsai, W. C., Tu, C. S., Hong, C. M., & Lin, W. M. (2023). A review of state-of-the-art and short-term forecasting models for solar PV power generation. Energies, 16(14), 5436.
Wang, G., Sun, S., Fan, S., Liu, Y., Cao, S., & Guan, R. (2024). A spatial–temporal data-driven deep learning framework for enhancing ultra-short-term prediction of distributed photovoltaic power generation. International Journal of Electrical Power & Energy Systems, 160, 110125.
Xiong, B., Chen, Y., Zhao, X., Su, Z., Fu, J., Chen, D., & Zhang, D. (2025). Multimodal ultra-short-term probabilistic solar power forecasting with generative AI and transformer. Advances in Applied Energy, 100250.
Yang, M., & Huang, X. (2018). Ultra-short-term prediction of photovoltaic power based on periodic extraction of PV energy and LSH algorithm. IEEE Access, 6, 51200-51205.
Ye, X., Yin, J., Zhang, J., Li, A., Liu, Z., Chen, B., ... & Li, H. (2026). A Multi-Scale CNN-Transformer Network with Residual Correction for Ultra-Short-Term Photovoltaic Power Forecasting. Processes, 14(5), 759.
Yong, B., Zhang, Y., Shen, J., Ren, A., Zhou, X., & Zhou, Q. (2025). ConvODE-Mixer: A multimodal deep learning model for ultra-short-term PV power forecasting. Solar Energy, 300, 113777.
Zazoum, B. (2022). Solar photovoltaic power prediction using different machine learning methods. Energy Reports, 8, 19-25.
Zhai, C., He, X., Cao, Z., Abdou-Tankari, M., Wang, Y., & Zhang, M. (2025). Photovoltaic power forecasting based on VMD-SSA-Transformer: Multidimensional analysis of dataset length, weather mutation and forecast accuracy. Energy, 324, 135971.
Zhang, J., Zhao, Z., Guo, R., Hu, X., Qu, T., Ge, C., & Yan, J. (2026). A Novel Ultra-Short-Term PV Power Forecasting Method Based on a Temporal Attention-Variable Parallel Fusion Encoder Network. Energies, 19(1), 274.
Zhang, R., Wu, Y., Zhang, L., Xu, C., Wang, Z., Zhang, Y., ... & Chen, Q. (2025). A multiscale network with mixed features and extended regional weather forecasts for predicting short-term photovoltaic power. Energy, 318, 134792.
Zhou, D., Liu, Y., Wang, X., Wang, F., & Jia, Y. (2025). Combined ultra-short-term photovoltaic power prediction based on CEEMDAN decomposition and RIME optimized AM-TCN-BiLSTM. Energy, 318, 134847.
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Zhengqi Fu (Author)
This work is licensed under a Creative Commons Attribution 4.0 International License.
