Adaptive Graph Neural Network-Based Hybrid Approach for Long-Term Photovoltaic Power Forecasting
Source abstract
Abstract
Photovoltaic power generation prediction is crucial for the effective integration of renewable energy into the grid, real-time grid balancing, and the optimization of energy storage systems. However, PV power generation is highly dependent on environmental factors such as weather conditions. Photovoltaic power generation prediction is crucial for the effective integration of renewable energy into the grid, real-time grid balancing, and the optimization of energy storage systems. However, PV power generation is highly dependent on environmental factors such as weather conditions. Effectively integrating environmental information remains a major challenge for photovoltaic power forecasting. This study proposes a hybrid deep learning model that incorporates an adaptive neural network to capture the latent relationships between PV power generation and environmental variables, thereby enhancing forecasting accuracy. The adaptive graph neural network employs a data-driven directed graph structure, where TCN and variable interaction layers are alternately stacked to better model the spatiotemporal coupling among variables for long-term PV output forecasting. The proposed model was evaluated on three sites located in different regions, with a fixed input length of 96 and output horizons ranging from 96 to 768 steps. Compared with state-of-the-art baselines, the model achieved average improvements of 2.19% and 1.57% in MSE and MAE at a 384-step horizon, and 2.81% and 2.47% at a 768-step horizon, respectively, demonstrating superior performance in long-term PV output forecasting tasks.
Source: crossref · Quality: baseline
Research context
OpenAlex topics: Solar Radiation and Photovoltaics · Energy Load and Power Forecasting · Photovoltaic System Optimization Techniques
SciVX discovery tags are terminology-based. OpenAlex topics are source-provided classifications.
Affiliations
College of Mechanical and Vehicle Engineering, Hunan University, Changsha 410082, China
College of Art and Science, The Ohio State University, Columbus, OH 43210, USA
Funding
National Science Foundation of China Excellent Young Overseas Talents Programme · Science and Technology Innovation Program (Overseas Expert Program) of Hunan Province
References (44)
Show source-provided references
1. Novel Grass Hopper optimization based MPPT of PV systems for complex partial shading conditions10.1016/j.solener.2020.01.070 ↗
2. Multi-step prediction of photovoltaic power based on two-stage decomposition and BILSTM10.1016/j.neucom.2022.06.117 ↗
3. Potential reductions of CO2 emissions from the transition to electric vehicles: Thailand’s scenarios towards 203010.1016/j.egyr.2023.08.073 ↗
4. Cost-benefit analysis of remote hybrid wind-diesel power stations: Case study Aegean Sea islands10.1016/j.enpol.2006.04.012 ↗
5. Generative adversarial networks and convolutional neural networks based weather classification model for day ahead short-term photovoltaic power forecasting10.1016/j.enconman.2018.11.074 ↗
6. COA-CNN-LSTM: Coati optimization algorithm-based hybrid deep learning model for PV/wind power forecasting in smart grid applications10.1016/j.apenergy.2023.121638 ↗
7. Forecasting of photovoltaic power generation and model optimization: A review10.1016/j.rser.2017.08.017 ↗
8. Extensive comparison of physical models for photovoltaic power forecasting10.1016/j.apenergy.2020.116239 ↗
9. A review and evaluation of the state-of-the-art in PV solar power forecasting: Techniques and optimization10.1016/j.rser.2020.109792 ↗
10. Photovoltaic power forecasting model employing epoch-dependent adaptive loss weighting and data assimilation10.1016/j.solener.2025.113351 ↗
11. The comparison of Holt–Winters method and Multiple regression method: A case study10.1016/j.energy.2016.04.115 ↗
12. Short-term reforecasting of power output from a 48 MWe solar PV plant10.1016/j.solener.2014.11.017 ↗
13. A hybrid model (SARIMA–SVM) for short-term power forecasting of a small-scale grid-connected photovoltaic plant10.1016/j.solener.2013.10.002 ↗
14. Kushwaha, V., and Pindoriya, N.M. (2017, January 21–23). Very short-term solar PV generation forecast using SARIMA model: A case study. Proceedings of the 2017 7th International Conference on Power Systems (ICPS), Pune, India.10.1109/ICPES.2017.8387332 ↗
15. Solar irradiance feature extraction and support vector machines based weather status pattern recognition model for short-term photovoltaic power forecasting10.1016/j.enbuild.2014.10.002 ↗
16. Time series forecasting for hourly photovoltaic power using conditional generative adversarial network and Bi-LSTM10.1016/j.energy.2022.123403 ↗
17. Accurate solar PV power prediction interval method based on frequency-domain decomposition and LSTM model10.1016/j.energy.2022.125592 ↗
18. Short-term self consumption PV plant power production forecasts based on hybrid CNN-LSTM, ConvLSTM models10.1016/j.renene.2021.05.095 ↗
19. Lea, C., Vidal, R., Reiter, A., and Hager, G.D. (2016, January 11–14). Temporal convolutional networks: A unified approach to action segmentation. Proceedings of the European Conference on Computer Vision, Amsterdam, The Netherlands.10.1007/978-3-319-49409-8_7 ↗
20. Ultra-short term wind power prediction applying a novel model named SATCN-LSTM10.1016/j.enconman.2021.115036 ↗
21. ImageNet classification with deep convolutional neural networks10.1145/3065386 ↗
22. Convolutional neural networks for time series classification10.21629/JSEE.2017.01.18 ↗
23. Recurrent neural networks and robust time series prediction10.1109/72.279188 ↗
24. Wen, Q., Zhou, T., Zhang, C., Chen, W., Ma, Z., Yan, J., and Sun, L. (2023, January 19–25). Transformers in time series: A survey. Proceedings of the Thirty-Second International Joint Conference on Artificial Intelligence, Macao.10.24963/ijcai.2023/759 ↗
25. Spatio-temporal graph neural networks for predictive learning in urban computing: A survey10.1109/TKDE.2023.3333824 ↗
26. Qin, Y., Song, D., Chen, H., Cheng, W., Jiang, G., and Cottrell, G. (2017). A dual-stage attention-based recurrent neural network for time series prediction. arXiv.10.24963/ijcai.2017/366 ↗
27. Deep learning for spatio-temporal data mining: A survey10.1109/TKDE.2020.3025580 ↗
28. Data-driven intelligent transportation systems: A survey10.1109/TITS.2011.2158001 ↗
29. A graph attention network framework for generalized-horizon multi-plant solar power generation forecasting using heterogeneous data10.1016/j.renene.2025.122520 ↗
30. Ultra-short-term distributed PV power forecasting for virtual power plant considering data-scarce scenarios10.1016/j.apenergy.2024.123890 ↗
31. Wang, Y., Zhao, Z., Wang, F., Sun, S., Cheng, Y., and Yu, J. (IEEE Trans. Ind. Appl., 2025). A Distributed PV Ultra-short-term Power Deterministic and Probabilistic Forecasting Based on Dynamic Graph Network with Shape-amplitude Criteria, IEEE Trans. Ind. Appl., in press.10.1109/TIA.2025.3604364 ↗
32. Short-term Probabilistic Forecasting for Regional PV Power based on Convolutional Graph Neural Network and Parameter Transferring10.1109/TPWRS.2024.3503288 ↗
33. Wang, Y., Dai, H., Zhen, Z., Wang, F., Wang, J., and Lu, L. (2024, January 21–25). Domain Adversarial Graph Neural Network based Regional Distributed PV Ultra-Short-term Power Forecasting Considering Data-Scarce Scenarios. Proceedings of the 2024 IEEE Power Energy Society General Meeting (PESGM), Seattle, WA, USA.10.1109/PESGM51994.2024.10689089 ↗
34. Yang, M., Zhou, M., Kalander, M., Huang, Z., and King, I. (2021, January 14–18). Discrete-time temporal network embedding via implicit hierarchical learning in hyperbolic space. Proceedings of the 27th ACM SIGKDD Conference on Knowledge Discovery Data Mining, Singapore.10.1145/3447548.3467422 ↗
35. Novel STAttention GraphWaveNet model for residential household appliance prediction and energy structure optimization10.1016/j.energy.2024.132582 ↗
36. An embedded spatiotemporal hybrid model integrating multi-graphs and attention-driven fusion for single-and multi-site photovoltaic power forecasting10.1016/j.enconman.2025.119897 ↗
37. Wu, Z., Pan, S., Long, G., Jiang, J., Chang, X., and Zhang, C. (2020, January 23–27). Connecting the dots: Multivariate time series forecasting with graph neural networks. Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery Data Mining, Virtual.10.1145/3394486.3403118 ↗
38. Zhang, T., Zhang, Y., Cao, W., Bian, J., Yi, X., Zheng, S., and Li, J. (2022). Less is more: Fast multivariate time series forecasting with light sampling-oriented mlp structures. arXiv.
39. Solar and wind power data from the Chinese state grid renewable energy generation forecasting competition10.1038/s41597-022-01696-6 ↗
40. Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A.N., Kaiser, Ł., and Polosukhin, I. (2017). Attention is all you need. Adv. Neural Inf. Process. Syst., 30.
41. Autoformer: Decomposition transformers with auto-correlation for long-term series forecasting
42. Zhou, H., Zhang, S., Peng, J., Zhang, S., Li, J., Xiong, H., and Zhang, W. (2021, January 19–21). Informer: Beyond efficient transformer for long sequence time-series forecasting. Proceedings of the AAAI Conference on Artificial Intelligence, Virtual. No. 12.10.1609/aaai.v35i12.17325 ↗
43. Zhang, Y., and Yan, J. (2023, January 1–5). Crossformer: Transformer utilizing cross-dimension dependency for multivariate time series forecasting. Proceedings of the Eleventh International Conference on Learning Representations, Kigali, Rwanda.
44. Wu, Z., Pan, S., Long, G., Jiang, J., and Zhang, C. (2019). Graph wavenet for deep spatial-temporal graph modeling. arXiv.10.24963/ijcai.2019/264 ↗