Securing Power Control Systems from Cyber Attacks:
Development of a Novel High Precision Detection Technique for Stable Control

Under the JST CREST Program, Project Assistant Professor Yacine Chakhchoukh and Associate Professor Hideaki Ishii of Tokyo Institute of Technology have jointly developed a novel technique for significantly improving robustness against cyber attacks via data falsifications targeting the socalled state estimation systems, which is one of the basic components for the monitoring and control of transmission power systems.

In future power systems, it is expected that distributed generation such as PVs and power generation systems will be introduced in large quantities, and various companies and organizations will take part in the operation. On the other hand, to accurately measure and estimate the states of the power grid in real-time, data regarding generated power, consumed power and so on must be communicated over networks. Such communication is likely to become targets of cyber attacks, which may result in serious physical accidents and even blackouts if control and measurement data is falsified by malicious attackers with sufficient knowledge of the system. Cyber security for power grids has recently emerged as an critical issue at the industry and government levels.

In this research, Profs. Chakhchoukh and Ishii have developed a novel algorithm to carry out state estimation in a stable manner even in the presence of highly malicious attacks. The algorithm employs statistical estimation methods robust against outliers in measurement noises and is capable of detecting the attacked parts in the system. Moreover, they have demonstrated through detailed simulation studies the advantages of the proposed algorithm in comparison with conventional methods. The results exhibited outstanding performance in detecting cyber attacks against transmission power grids and in particular in enabling stable state estimation even in the presence of falsification of the grid data stored in the energy management systems. The group at Tokyo Tech will extend their work by considering more realistic transmission grids, which may eventually lead them to apply to actual systems.

The outcome of this research will become publically available through the online version of the IEEE Transactions on Power Systems published by the Institute of Electric and Electronic Engineers (IEEE) on February 4th, 2016 (US Eastern Time).