An Integrated Cyber-physical Security Model for a Decentralized and Autonomous Smart Power Gri

There have been significant advancements in research and development of autonomous technologies to provide self-managing capabilities for complex systems such as the power grid in recent years. The power grid needs to be modernized to be able to self-heal from power failure, self-configure to minimize cost and optimize power flow and automate other power systems management functions in response to the changing energy demands and complex behavior of systems that rely on the power grid to operate. The smart grid is the next generation power grid being designed to provide these functions while ensuring the security and integrity of processes and operations in the traditional power grid.;The power grid consists of subsystems and components that are widely distributed, covering large geographic areas. To realize the goals of the smart grid, more information, communication, and digital technology would have to be integrated to facilitate information exchanges between the distributed components. The heavy reliance of the smart grid on communications network will open up the power system even more to the possibility of cyber-attacks from internal and external adversaries to the smart gird. Therefore, the smart grid should be designed to be reasonably secure from cyber-attacks and resilient to the impact of cyber-attacks and physical disturbance to the power grid infrastructure.;This dissertation demonstrates how to meet the cyber-security and resiliency requirements of the smart grid by developing a cyber-physical security model that integrates the physics of power transmission and the behavior of power automation functions into traditional network security controls to enable a more robust and dynamic cyber-security solution for the power grid. In this dissertation, I discuss my novel contributions that focus on the key requirements for resilient security in the complex and dynamic smart grid comprising: (1) A decentralized power system control and communications architecture that uses the power transmission line characteristics to decouple the grid into subsystems; (2) A communications protocol suite for decentralized autonomous networks which easily integrates into any layer of the TCP/IP protocol stack; and (3) A cyber-physical security model that takes advantage of the double-coupling characteristics of our proposed system architecture to precisely identify anomaly in the communications network. In this work, I also present models and algorithms for decentralized control and intrusion detection. I used these algorithms to achieve self-healing, critical load management, and interlocking automation functions for the smart grid.