Anomaly detection in power system control center critical infrastructures using rough classification algorithm

Power system control centers are moving toward distributed and decentralized operations. The use of information technology (IT) to achieve these goals produces vulnerabilities and security threats. To safeguard against the threat of cyber-attacks, service providers also need to maintain the accuracy, assurance and integrity of their interdependent data networks. This paper presents the results of the anomaly detection technique using rough sets classification algorithm for improving the security of power system control centers in the electric power system critical infrastructure. The methodology presented can be used to identify attacks and failures and, also, for improving the confidence of the state estimation process. A test environment is implemented and the results for a 6 bus electrical power system are presented.