Intelligent False Data Injection Attack Detection Using Soft Computing in Cyber-Physical Power Systems

The modern power system has developed into a cyber-physical system with the incorporation of smart meters, Phasor Measurement Units (PMUs), extensive area monitoring, control systems and substation automation. False Data Injection Attack (FDIA) has gained popularity in Cyber-Physical Power System (CPPS). In particular, there is an increasing need to examine the possible behavior of adversaries to enhance the cybersecurity nature of CPPS but the FDIA model is frequently built on the effects of attacker or the effect of attack. Therefore, this study presents an Intelligent False Data Injection Attack Detection using Soft Computing (IFDIAD-SD) technique in the CPPS environment. The aim of the IFDIAD-SD technique is the effectual recognition and classification of FDIAs in the CPPS environment. In the IFDIAD-SD method, fuzzy min-max normalization is used to scale the input data. For FDIA detection and classification, convolutional neural network with autoencoder (CNN-AE) method is applied. At final stage, the multiverse optimization (MVO) technique is used to optimally select the parameter values of the CNN-AE model. A series of experiments were implemented for identifying the superior performance of the IFDIAD-SD technique. The comprehensive comparative result analysis revealed the supremacy of the IFDIAD-SD method when compared to other models.