Analysis of IIoT data using visibility graphs and complex network methods
Frontiers in Physics · 29.06.2026 · 14
DOI: 10.3389/fphy.2026.1864040 ↗Kurz gefasst
Fachartikel in der Zeitschrift Frontiers in Physics, online erschienen am 29.06.2026. Verfasst von Jiwei Xu und 5 weiteren Autor:innen. Die Arbeit nennt 31 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0).
Abstract (Originalsprache)
With the increasing volume of time series data generated in the Industrial Internet of Things (IIoT), characterizing the intrinsic structural relationships within such data has become a key challenge in complex system analysis. The visibility graph (VG) method maps time series into complex networks, enabling the representation of dynamic features in a topological form and providing an effective tool for nonlinear time series analysis. Compared with traditional approaches that primarily focus on statistical features, VG can reveal latent structural information and correlation patterns within the sequence. Based on this, the VG method is applied in this study to the analysis of power consumption time series from electric vehicle charging stations. The time series are transformed into complex networks, and their structural characteristics are systematically investigated from a complex network perspective. By constructing visibility graph networks of the power consumption sequences, statistical analyses of node degree, degree distribution, and overall topological structure are performed to characterize differences in system structure under different operating states and scenarios. The results show that the constructed VG networks exhibit pronounced structural heterogeneity. The node degree distributions display clear heavy-tailed behavior and approximately follow a power-law scaling within a certain range. Meanwhile, a small number of highly connected nodes play a dominant role in the overall network structure, while most nodes remain weakly connected. Further analysis reveals significant structural differences across operating conditions, reflecting the complex dynamic evolution of charging behavior under different scenarios. On this basis, classification models are built by combining statistical features with VG-based structural features for validation. The results demonstrate that incorporating VG features can improve classification performance to a certain extent, indicating that the VG method provides additional structural information for time series analysis. Overall, the proposed approach offers a useful framework for structural modeling and anomaly detection of IIoT time series data.
Zitieren
APA
Xu, J., Wang, K., Wang, J., Wu, J., Lv, X., & Song, Q. (2026). Analysis of IIoT data using visibility graphs and complex network methods. Frontiers in Physics, 14. https://doi.org/10.3389/fphy.2026.1864040
Vancouver
Xu J, Wang K, Wang J, Wu J, Lv X, Song Q. Analysis of IIoT data using visibility graphs and complex network methods. Frontiers in Physics. 2026;14. doi:10.3389/fphy.2026.1864040
BibTeX
@article{Xu2026,
author = {Xu, Jiwei and Wang, Kaigong and Wang, Jiaqi and Wu, Jiayu and Lv, Xinyan and Song, Qun},
title = {Analysis of IIoT data using visibility graphs and complex network methods},
journal = {Frontiers in Physics},
year = {2026},
volume = {14},
doi = {10.3389/fphy.2026.1864040}
}
EUVIMED