FDIA Localization Detection Based on Intuitionistic Fuzzy Set-multiple Hidden Layer Random Vector Functional Link Network

False data injection attacks seriously threaten the security and stability of the cyber-physical power system. To address the issue that existing attack detection methods struggle to accurately identify and locate the attacked bus, a location detection method based on an intuitionistic fuzzy set and a multiple hidden-layer random vector function link network is proposed in this paper. The proposed method extracts the data feature information of the complete measurement through a random vector function connection network to reduce the missed detection rate and expands it into a multi-hidden layer architecture to prevent overfitting of the detection method effectively. Meanwhile, the intuitionistic fuzzy method is employed to mitigate the impact of measurement noise on detection accuracy, thereby achieving high-precision attack location detection. A large number of experiments are conducted on the test systems of IEEE-14, IEEE-57, and IEEE-118 buses to verify the effectiveness of the proposed method. Compared with various methods, it is verified that the proposed method has better accuracy, precision, F1 value, recall rate and AUC value.