Skip to main content
U.S. Department of Energy
Office of Scientific and Technical Information

Connectivity, Centrality, and Bottleneckedness: On Graph Theoretic Methods for Power Systems

Technical Report ·
DOI:https://doi.org/10.2172/3012328· OSTI ID:3012328

Grid Architecture is largely concerned with structures since we view the power grid as being comprised of a network of structures: electrical infrastructure, industry structure control structure, digital superstructure including communications networks, and convergent structures such as those for water, gas, and transport. These structures are interconnected in complex ways, and many of the characteristics of the grid that we wish to enhance or add derive directly from structure or are strongly influenced by it. Consequently it is important to have rigorous methods to analyze such structures and systematically modify them. These structures are far too complex to handled by inspection and hence the focus here on applying graph theory to grid structure problems. Graphs are abstract structures that express pairwise relationships between entities. Because of their versatility and universality, graphs are a natural data structure for representing a myriad of complex systems. The burgeoning field of network science, for which graph theory serves as a mathematical scaffold, attests to the ubiquity and utility of graph theoretic analyses in farranging disciplines, including biology, chemistry, social science, and engineering [2]. In the case of power systems, the application of graph theoretic methods is far from new; see [12] for a survey of scientific literature on graph theoretic methods applied to the electric grid. Rather than provide a comprehensive survey here, we aim to provide a self-contained introduction to selected graph theoretic topics that may have increased pertinence in the structural analysis of power systems. In particular, we focus on methodologies for defining, scoring, and identifying connectivity, centrality, and bottleneckeness properties in graphs. We apply these measures to power systems graph data when possible, frequently present visualizations, and discuss the computational feasibility of these methods.

Research Organization:
Pacific Northwest National Laboratory (PNNL), Richland, WA (United States)
Sponsoring Organization:
USDOE
DOE Contract Number:
AC05-76RL01830
OSTI ID:
3012328
Report Number(s):
PNNL-29662
Country of Publication:
United States
Language:
English

Similar Records

Discrete Graph Structure Learning for Forecasting Multiple Time Series
Conference · Tue May 04 00:00:00 EDT 2021 · OSTI ID:1866733

A Transductive Graph Neural Network learning for Grid Resilience Analysis
Conference · Tue Oct 31 00:00:00 EDT 2023 · OSTI ID:2998563

Disordered topological graphs enhancing nonlinear phenomena
Journal Article · Wed Apr 05 00:00:00 EDT 2023 · Science Advances · OSTI ID:2470864

Related Subjects