Logo image
Back
Intrusion Detection in Smart Grids Using Deep Belief Network and K-Means Clustering Model for Electric Vehicle Charging Systems
Conference paper

Intrusion Detection in Smart Grids Using Deep Belief Network and K-Means Clustering Model for Electric Vehicle Charging Systems

P. Chandra Sekhar, Murali Karri, S. Sree Vidhya, Anil Lokesh Gadi, Sree Lakshmi Lingineni and Anvesh Perada
2025 5th International Conference on Emerging Research in Electronics, Computer Science and Technology (ICERECT), pp 1-6
12 Sep 2025
DOI: https://doi.org/10.1109/ICERECT65215.2025.11379797

Abstract

Accuracy Binary Cuckoo Search (BCS) Biological system modeling Computational modeling Deep Belief Networks (DBN) Density Peak Clustering Algorithm (DPCA) Electric vehicle (EV) Encoding Feature extraction Intrusion detection Intrusion Detection (ID) Intrusion Detection System (IDS) Smart Grid (SG) Smart grids Training User experience Internet of Things
The safety and reliability of today's electricity grids depend on intrusion detection in smart grids. Intelligent detection systems that can accurately identify possible risks are urgently needed due to the increasing volume and complexity of smart grid data. To guarantee high-quality data input for analysis, this study offers a new method that starts with data pretreatment, which includes cleaning and encoding. The Binary Cuckoo Search algorithm is used for feature selection in order to identify the most relevant attributes. As a follow-up, it provides an MDPCA-and DBN based fuzzy aggregation approach. To reduce dataset size and fix class imbalance, the MDPCA method divides the training dataset into subsets with comparable attribute distributions. To achieve more precise training, a distinct sub-DBN classifier is trained for each subset. Outperforming multiple other models examined in this research, the proposed model known as Modified Density PCA-DBN achieves an outstanding accuracy of 95.84%. The efficacy of integrating MDPCA and DBN inside a fuzzy aggregation architecture for strong intrusion detection is proven by these outcomes. By facilitating more precise and extensible threat detection procedures, this paradigm provides a viable approach to improving cybersecurity in smart grid systems.

Metrics

1 Record Views

Details

Logo image