Risk Assessment of High-Voltage Power Grid Under Typhoon Disaster Based on Model-Driven and Data-Driven Methods

Xiao Zhou and Jiang Li ()
Additional contact information
Xiao Zhou: College of Electrical Engineering, Shanghai University of Electric Power, Shanghai 200090, China
Jiang Li: College of Electrical Engineering, Shanghai University of Electric Power, Shanghai 200090, China

Energies, 2025, vol. 18, issue 4, 1-18

Abstract: As global warming continues to intensify, typhoon disasters will more frequently occur in East and Southeast Asia, posing a high risk of causing large-scale power outages in the power system. To investigate the impact of typhoon disasters on high-voltage power grids, a comprehensive risk assessment method that integrates model-driven and data-driven approaches is proposed, which can predict power grid faults in advance and provide support for power grid operators to generate emergency dispatching plans. Firstly, by comparing actual loads with the design strengths of the transmission tower-line system and analyzing the geometric relationship between typhoon wind circles and the system, key variables, such as wind speed, longitude, latitude, and other pertinent factors, are screened. The Spearman correlation coefficient is employed to pinpoint the meteorological variables that exhibit a high degree of relevance, enhancing the accuracy and interpretability of our model. Secondly, addressing the lack of power grid fault samples, three data balancing methods—Borderline-SMOTE, ADASYN, and SMOTE-Tomek—are compared, with Borderline-SMOTE selected for its superior performance in enhancing the sample set. Additionally, a power grid failure risk assessment model is built based on Light Gradient Boosting Machine (LightGBM), and the Borderline-Smoothing Algorithm (BSA) is used for the modeling of power grid faults. The nonlinear mapping relationship between typhoon meteorological data and the power grid equipment failure rate is extracted through deep learning training. Subsequently, the Tree-structured Parzen Estimator (TPE) is leveraged to optimize the hyperparameters of the LightGBM model, thus enhancing its prediction accuracy. Finally, the actual power system data of a province in China under a strong typhoon are assessed, validating the proposed assessment method’s effectiveness.

Keywords: typhoon disaster; high-voltage power grid; risk assessment; failure probability; deep learning (search for similar items in EconPapers)
JEL-codes: Q Q0 Q4 Q40 Q41 Q42 Q43 Q47 Q48 Q49 (search for similar items in EconPapers)
Date: 2025
References: View references in EconPapers View complete reference list from CitEc
Citations:

Downloads: (external link)
https://www.mdpi.com/1996-1073/18/4/809/pdf (application/pdf)
https://www.mdpi.com/1996-1073/18/4/809/ (text/html)

Related works:
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:gam:jeners:v:18:y:2025:i:4:p:809-:d:1587213

Access Statistics for this article

Energies is currently edited by Ms. Agatha Cao

More articles in Energies from MDPI
Bibliographic data for series maintained by MDPI Indexing Manager ().