Intrusion Detection and Prevention at the End-User Level in Smart Grid Cyber-Physical Systems

In order to keep the electric power system running smoothly and reliably, it is constantly monitored and controlled by an intelligent cyber layer that consists of data processing algorithms and a wide network of sensors. Cyberattacks on this data processing and collecting system might halt operations and have catastrophic physical consequences like system breakdowns. A common kind of attack is the FDIA. To avoid detection by the state estimator's inherent anomaly detection capabilities, an attacker tries to trick the grid's underlying control system into producing disturbances. On top of that, there are major issues with customer-end security. For instance, criminals might intercept, alter, or replay wireless transmissions of power consumption signals sent by consumers to their utility provider. Consequently, smart grid security and privacy are major apprehensions. In order to prevent cyberattacks on the vital infrastructures of the electrical system, this thesis helps. New detection systems for the generating and transmission side, as well as the end-user customer side infrastructure, are the primary focus of the study. To protect the generating and transmission sides from cyber-attack, we provide a system based on intelligent sensor weights and an optimization strategy based on sophisticated grey wolves. In addition, to safeguard end-user customer data from cyberattacks, we provide a gaussian mixture model–based privacy protection strategy and an artificial neural structure–based intelligent loop detection method.