Comparative Assessment of Machine Learning (ML) Techniques for False Data Injection Attack (FDIA) Classification

With the advancement in communication technologies, the cyber physical systems have become vulnerable to various cyber threats. In this paper, one such significant cyber threat in the microgrids, viz., False Data Injection Attack (FDIA), is modeled in various forms, and three popular machine learning methods have been employed to detect and classify them. The FDIA classification strategy thus developed is tested on the IEEE 13-bus Islanded AC Microgrid test system, constituted of four DGs of similar rating supplying multiple loads in the network, and is modelled in Real Time Digital Simulation (RTDS) platform.