Abstract: False data injection attacks can destroy the data integrity and availability in SCADA systems, so the detection method for data security and stable operation of smart grid is of great significance. The extended Kalman filter method uses the detected data to recursively obtain the real-time state of the system, so as to achieve the purpose of effectively detecting and checking whether or not there are false data injections in the power system. In addition, this method can not only evaluate the past operating states, but also predict the minute changes of future system. Furthermore, this method can effectively identify the false data injection attack under unknown conditions of the system. IEEE-14 node and 30-node models are used as the benchmarks for detecting false data injection attacks. The results show that the proposed method not only is superior to traditional state estimation methods that fail to detect false data injection attack, but also exhibits many advantages, such as small storage capacity and easy realization.

Key words: false data injection attack, smart grid, extended Kalman filter, cyber attack