Locational Detection of False Data Injection Attack in Smart Grid Based on Multilabel Machine Learning Classification Methods

State estimation is important for the smart grid monitoring and control. Owing to information and communication technologies, smart grid performs efficient management of power system reliability and stability. However, these information and communication technologies also make smart grid vulnerable to cyber threats, particularly false data injection attacks (FDIAs). It is the most threatening attack as it disrupts the operations of the grid and manipulates the reading of the state estimator. In smart grid systems, power measurements are obtained through various advanced metering systems and the location detection of compromised meters is also important besides determining the FDIA attack. This paper propose multilabel machine learning classification methods, binary relevance and classifier chain, to detect FDIA and locate compromised smart meters. Through a comprehensive experiment on IEEE 14 bus system, we showed that the accuracy of binary relevance is 95.1%.