Next Article in Journal
Effects of La2O3/Al2O3 Ratio in Slag Used for Electroslag Remelting on La Yield and Cleanliness of H13 Steel
Previous Article in Journal
Functional and Bioactive Characterization of Hemp Cake Proteins and Polyphenols from Non-Psychoactive Cannabis sativa
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
This is an early access version, the complete PDF, HTML, and XML versions will be available soon.
Article

An End-to-End Framework for Spatiotemporal Data Recovery and Unsupervised Cluster Partitioning in Distributed PV Systems

1
State Grid Jibei Electric Power Company Limited, Beijing 100032, China
2
School of Electrical Engineering, Southeast University, Nanjing 210096, China
*
Author to whom correspondence should be addressed.
Processes 2025, 13(10), 3186; https://doi.org/10.3390/pr13103186
Submission received: 16 September 2025 / Revised: 2 October 2025 / Accepted: 4 October 2025 / Published: 7 October 2025
(This article belongs to the Section Energy Systems)

Abstract

The growing penetration of distributed photovoltaic (PV) systems presents significant operational challenges for power grids, driven by the scarcity of historical data and the high spatiotemporal variability of PV generation. To address these challenges, we propose Generative Reconstruction and Adaptive Identification via Latents (GRAIL), a unified, end-to-end framework that integrates generative modeling with adaptive clustering to discover latent structures and representative scenarios in PV datasets. GRAIL operates through a closed-loop mechanism where clustering feedback guides a cluster-aware data generation process, and the resulting generative augmentation strengthens partitioning in the latent space. Evaluated on a real-world, multi-site PV dataset with a high missing data rate of 45.4%, GRAIL consistently outperforms both classical clustering algorithms and deep embedding-based methods. Specifically, GRAIL achieves a Silhouette Score of 0.969, a Calinski–Harabasz index exceeding 4.132×106, and a Davies–Bouldin index of 0.042, demonstrating superior intra-cluster compactness and inter-cluster separation. The framework also yields a normalized entropy of 0.994, which indicates highly balanced partitioning. These results underscore that coupling data generation with clustering is a powerful strategy for expressive and robust structure learning in data-sparse environments. Notably, GRAIL achieves significant performance gains over the strongest deep learning baseline that lacks a generative component, securing the highest composite score among all evaluated methods. The framework is also computationally efficient. Its alternating optimization converges rapidly, and clustering and reconstruction metrics stabilize within approximately six iterations. Beyond quantitative performance, GRAIL produces physically interpretable clusters that correspond to distinct weather-driven regimes and capture cross-site dependencies. These clusters serve as compact and robust state descriptors, valuable for downstream applications such as PV forecasting, dispatch optimization, and intelligent energy management in modern power systems.
Keywords: distributed photovoltaic systems; generative modeling; unsupervised clustering; deep representation learning; joint optimization; LSTM; autoencoder; HDBSCAN distributed photovoltaic systems; generative modeling; unsupervised clustering; deep representation learning; joint optimization; LSTM; autoencoder; HDBSCAN

Share and Cite

MDPI and ACS Style

Zhai, B.; Li, Y.; Qiu, W.; Zhang, R.; Jiang, Z.; Zeng, Y.; Qian, T.; Hu, Q. An End-to-End Framework for Spatiotemporal Data Recovery and Unsupervised Cluster Partitioning in Distributed PV Systems. Processes 2025, 13, 3186. https://doi.org/10.3390/pr13103186

AMA Style

Zhai B, Li Y, Qiu W, Zhang R, Jiang Z, Zeng Y, Qian T, Hu Q. An End-to-End Framework for Spatiotemporal Data Recovery and Unsupervised Cluster Partitioning in Distributed PV Systems. Processes. 2025; 13(10):3186. https://doi.org/10.3390/pr13103186

Chicago/Turabian Style

Zhai, Bingxu, Yuanzhuo Li, Wei Qiu, Rui Zhang, Zhilin Jiang, Yinuo Zeng, Tao Qian, and Qinran Hu. 2025. "An End-to-End Framework for Spatiotemporal Data Recovery and Unsupervised Cluster Partitioning in Distributed PV Systems" Processes 13, no. 10: 3186. https://doi.org/10.3390/pr13103186

APA Style

Zhai, B., Li, Y., Qiu, W., Zhang, R., Jiang, Z., Zeng, Y., Qian, T., & Hu, Q. (2025). An End-to-End Framework for Spatiotemporal Data Recovery and Unsupervised Cluster Partitioning in Distributed PV Systems. Processes, 13(10), 3186. https://doi.org/10.3390/pr13103186

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop