Intrusion detection based on hybrid classifiers for smart grid. (July 2021)

• Record Type: Journal Article
• Authors: Song, Chunhe; Sun, Yingying; Han, Guangjie; Rodrigues, Joel J.P.C.
• Is Part Of: Computers & electrical engineering. Volume 93(2021)
• Publication Date: 2021-07
• Subjects: Smart grid -- Intrusion detection -- Deep learning -- LSTM -- XGBoost; Computer engineering -- Periodicals; Electrical engineering -- Periodicals; Electrical engineering -- Data processing -- Periodicals; Ordinateurs -- Conception et construction -- Périodiques; Électrotechnique -- Périodiques; Électrotechnique -- Informatique -- Périodiques; Computer engineering; Electrical engineering; Electrical engineering -- Data processing; Periodicals; Electronic journals; 621.302854
• DOI: 10.1016/j.compeleceng.2021.107212
• Languages: English
• ISSNs: 0045-7906

Abstract

Abstract: In this paper, a novel intrusion detection method combining a deep learning-based method and a feature-based method is proposed for smart grid. Specifically, long short-term memory and extreme gradient boosting are adopted for intrusion detection, and the results are fused based on the accuracies of these two models. As the XGBoost method is sensitive to its parameters and unsuitable selections greatly degrade its performance, in this paper, a Bayesian method is proposed to optimize these parameters. Moreover, a crossover scheme in a genetic algorithm is introduced to reduce the impact of falling into a local optimum of Bayesian optimization. Extensive experimental results show the effectiveness of the proposed algorithm. Graphical abstract: Highlights: An intrusion detection method is proposed combining two classifiers adaptively for smart grid. A Bayesian based method is proposed to optimize the parameters of eXtreme Gradient Boosting. A method is proposed to reduce the impact of falling into local optimum of Bayesian optimization.