Abstract
In this paper, we extend our previous research on uniting the Cognitive Dynamic Systems (CDS) and the Smart Grid (SG) by introducing Cognitive Risk Control (CRC). The CDS is a structured physical model and research tool inspired by certain features of the brain. The CRC is an advanced feature of the CDS that embodies the concept of predictive adaptation allowing it to bring risk under control in situations involving unexpected or abnormal uncertainty such as a cyber-attack. The False Data Injection (FDI) attack is a special class of cyber-attack targeting the SG that is able to bypass the traditional bad data detection techniques. Here we will demonstrate how the entropic state, which is the objective function of the CDS, is able to detect and bring FDI attacks under control under the action of CRC. Through Task-Switch control, the CDS is able to switch on a new executive with different set of actions that affects the system configuration to bring the risk under control during an attack. With the CDS acting as the supervisor of the SG, simulations are carried out on a 4 bus-system and IEEE 14-bus system to demonstrate the capability of CRC when faced with FDI attacks. The results show that this system has great potential for future SG systems.
Highlights
The Cognitive Dynamic System (CDS) is an organized physical model and research tool that is rooted in certain aspects of the brain
This paper is novel for the following reasons: i) This is the first time that we have been able to incorporate the Cognitive Risk Control (CRC) with the previous the structure referenced in our earlier paper to give rise to a new construct that is able to bring the problem of attack in the Smart Grid (SG) under control
We believe that the architecture proposed has great potential in handling the risks associated with attacks that the grid will face in the future as the networks become more interconnected
Summary
The Cognitive Dynamic System (CDS) is an organized physical model and research tool that is rooted in certain aspects of the brain. The rest of the paper is organized as follows: Section II covers briefly the basic concepts associated with state estimation, bad data detection and FDI attacks in the power grid. As shown by the mathematical proof, the residuals due to the attack and those due to normal conditions are considered the same For this reason, since the bad data detection techniques rely on statistical methods for the calculation of these residuals, they are unable to detect that the measurement vector has been maliciously falsified. Since the bad data detection techniques rely on statistical methods for the calculation of these residuals, they are unable to detect that the measurement vector has been maliciously falsified This results in the calculation of wrong system states that can in turn start a domino effect with disastrous consequences. This computational iteration uses the preceding a posteriori estimates to predict new a priori estimates
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
