AUTHOR=Yang Zhichun , Min Huaidong , Zhao Jie , Dong Xuzhu , Yang Fan , Wang Chenhao , Zhang Nan TITLE=Vulnerability analysis of power grid structure based on complex network theory JOURNAL=Frontiers in Energy Research VOLUME=12 YEAR=2025 URL=https://www.frontiersin.org/journals/energy-research/articles/10.3389/fenrg.2024.1498678 DOI=10.3389/fenrg.2024.1498678 ISSN=2296-598X ABSTRACT=Introduction

The key nodes of an intelligent distribution network significantly impact the reliability and stability of the distribution network’s operation. The failure of these key nodes can severely affect the safe operation of the distribution network. Therefore, vulnerability analysis of key nodes is particularly important.

Methods

This article proposes a comprehensive weighting method for evaluating indicators,combining the analytic hierarchy process (AHP) and entropy weighting method, while considering the structure and operational status of the power system grid. Key node structural evaluation indicators, such as node degree, node shrinkage centrality, and electric mediator, are established considering “significance” and “destructiveness.” State indicators are established based on the degree of impact of current, voltage, and load changes on the grid, as well as the uniformity of their distribution, including the improved current distribution entropy, voltage terre entropy, and three-phase state indicators of lost load. Subsequently, based on the AHP and entropy weighting method, a comprehensive weighting method is proposed to assign subjective and objective weights to the comprehensive evaluation indicators, obtaining the comprehensive weights of the indicators. Finally, the gray correlation degree is introduced to improve the ideal solution of the multi-objective decision making method, obtaining the criticality of the grid nodes and then identifying the critical nodes.

Results and Discussion

The example analysis presented in this article shows that the identified critical nodes of the power grid have a high degree of overlap with the identification results of different methods, can better identify edge nodes, and validate the effectiveness of the proposed evaluation indicators and key node identification methods.