PCA-Optimized SVM Framework for Detecting False Data Injection Attacks in Battery Energy Storage Systems

Abstract

False Data Injection Attacks (FDIAs) pose a significant threat to the reliability of Battery Energy Storage Systems (BESS) in smart grids. This paper proposes a PCA-optimized Support Vector Machine (SVM) detection framework specifically designed for BESS. First, we model the impact of FDIAs on BESS operations, demonstrating how such attacks can manipulate state-of-charge, power output, and grid frequency to disrupt system efficiency and stability. Key operational indicators, including voltage, current, and power, are used for an extensive feature extraction process, after which Principal Component Analysis (PCA) is applied to select the most informative features and improve classification efficiency. The proposed framework is evaluated through comprehensive simulations, achieving a detection accuracy of 98.5%, precision of 97.8%, recall of 98.2%, and an F1-score of 98.0%. The confusion matrix and ROC curve confirm the robustness of the model, showing minimal false positives and false negatives. These results demonstrate that the proposed PCA-optimized SVM framework provides an effective and computationally efficient solution for mitigating FDIAs in BESS and strengthens the security foundation of future smart grids.