Enhancing Electrical Network Vulnerability Assessment with Machine Learning and Deep Learning Techniques - Article 2

Aug 3, 2026

Upload by Violet

This research applies advanced machine learning to evaluate node vulnerability in power grid networks using the SciGRID and GridKit datasets with 479 nodes and 16,167 nodes, and 765 and 20,539 edges. Missing data are handled through K-nearest neighbor and median imputation, while centrality metrics are fused into a single comprehensive score to rank nodes into four vulnerability-relevant levels. Traditional learning models (XGBoost, SVM, Multilayer Perceptron) and Graph Neural Networks are benchmarked for their ability to identify critical nodes using features beyond centrality alone, addressing gaps in prior centrality-only analysis.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages24

Reading Time1 h 0 min

AccessFree to download

Introduction

Data, preprocessing, and vulnerability scoring

Traditional machine learning and GNN modeling

Benchmarking and critical node identification

Motivation and research gap

Northeast Journal of Complex Systems (NEJCS) | Volume 6
Number 1 Special Issue: NERCCS 2024 Papers | Article 2 | | --- | --- | | June 2024
Enhancing Electrical Network Vulnerability Assessment with Machine Learning and Deep Learning Techniques
M Mishkatur Rahman
North Dakota State University, [mmishkatur.rahman@ndsu.edu](mmishkatur.rahman@ndsu.edu)
Ayman Sajjad Akash
North Dakota State University, [ayman.akash@ndsu.edu](ayman.akash@ndsu.edu)
Harun Pirim
North Dakota State University, [harun.pirim@ndsu.edu](harun.pirim@ndsu.edu)
Chau Le
North Dakota State University, [chau.le@ndsu.edu](chau.le@ndsu.edu)
Trung Le
University of South Florida, [tqle@usf.edu](tqle@usf.edu)
See next page for additional authors
Follow this and additional works at: [https://orb.binghamton.edu/nejcs](https://orb.binghamton.edu/nejcs)
Part of the Operations Research, Systems Engineering and Industrial Engineering Commons | | Recommended Citation Rahman, M Mishkatur; Akash, Ayman Sajjad; Pirim, Harun; Le, Chau; Le, Trung; and Yadav, Om Prakash (2024) "Enhancing Electrical Network Vulnerability Assessment with Machine Learning and Deep Learning Techniques," Northeast Journal of Complex Systems (NEJCS): Vol. 6 : No. 1 , Article 2. DOI: 10.22191/nejcs/vol6/iss1/2 Available at: [https://orb.binghamton.edu/nejcs/vol6/iss1/2](https://orb.binghamton.edu/nejcs/vol6/iss1/2) This Article is brought to you for free and open access by The Open Repository @ Binghamton (The ORB) . It has been accepted for inclusion in Northeast Journal of Complex Systems (NEJCS) by an authorized editor of The Open Repository @ Binghamton (The ORB) . For more information, please [contact ORB@binghamton.edu](contact ORB@binghamton.edu). Enhancing Electrical Network Vulnerability Assessment with Machine Learning and Deep Learning Techniques Authors M Mishkatur Rahman, Ayman Sajjad Akash, Harun Pirim, Chau Le, Trung Le, and Om Prakash Yadav This article is available in Northeast Journal of Complex Systems (NEJCS): [https://orb.binghamton.edu/nejcs/vol6/](https://orb.binghamton.edu/nejcs/vol6/)[ ](https://orb.binghamton.edu/nejcs/vol6/)iss1/2 Enhancing Electrical Network Vulnerability Assessment with Machine Learning and Deep Learning Techniques M Mishkatur Rahman 1 , Ayman Sajjad Akash 1 , Harun Pirim 1∗ , Chau Le2 , Trung (Tim) Q. Le3 and Om Prakash Yadav4 1Dept. of Industrial and Manufacturing Engineering, North Dakota State University, Fargo, ND, USA 2Dept. of Civil, Construction and Environmental Engineering, North Dakota State University, Fargo, ND, USA 3Dept. of Industrial and Management Systems Engineering, University of South Florida, Tampa, FL, USA 4Dept. of Industrial and Systems Engineering, North Carolina A&T State University, Greensboro, NC, USA ∗ Corresponding Author: [harun.pirim@ndsu.edu](harun.pirim@ndsu.edu) Abstract This research utilizes advanced machine learning techniques to evaluate node vulnerability in power grid networks. Utilizing the SciGRID and GridKit datasets, consisting of 479, 16,167 nodes and 765, 20,539 edges respectively, the study employs K-nearest neighbor and median imputation methods to address missing data. Centrality metrics are integrated into a single comprehensive score for assessing node criticality, categorizing nodes into four centrality levels informative of vulnerability. This categorization informs the use of traditional machine learning (including XGBoost, SVM, Multilayer Perceptron) and Graph Neural Networks in the analysis. The study not only benchmarks the capabilities of these models in network analysis but also explores their potential in identifying critical nodes using features beyond centrality metrics alone, enhancing their applicability in real-world scenarios. The research addresses a significant gap in effectively assessing the vulnerability of electrical networks, marked by isolated use of traditional centrality metrics and a lack of

FAQs

What datasets and preprocessing steps are used for evaluating electrical network vulnerability?

The study uses the SciGRID and GridKit datasets and applies K-nearest neighbor and median imputation to address missing data.

How is node vulnerability quantified in the research?

Centrality metrics are combined into a single comprehensive score, which classifies nodes into four centrality levels linked to vulnerability and criticality.

Which modeling approaches are compared for vulnerability assessment?

The research benchmarks traditional machine learning models such as XGBoost, SVM, and Multilayer Perceptron against Graph Neural Networks.

Aug 3, 2026

Upload by Violet