FDI Attacks Detection and Reliable Prediction for Lithium-Ion Battery Based on Encrypted Koopman Prediction Method

This paper proposes a method for detecting attacks and reliable prediction based on encrypted Koopman prediction method in the scenario where batteries are subject to FDI (False Data Injection) attacks. On the one hand, this method accurately detects whether the lithium-ion battery system has been subjected to FDI attacks. On the other hand, even in the event of FDI attacks, reliable prediction of the lithium-ion battery system output is achieved through the use of encryption and pre-set signal methods. The paper provides ample numerical comparisons to demonstrate the effectiveness of these results.