Resilient Distributed Coordination Control of Multiarea Power Systems Under Hybrid Attacks
Abstract : Resilient distributed coordination control is studied on multiarea power systems with low inertia under hybrid attacks, including denial-of-service (DoS) attack and deception attack. The communication among various areas under the DoS attack is deteriorated to switching residual topologies whose time characteristic is modeled by model-dependent average dwell time (MDADT). Deception attack with malicious strategy targeting at negative feedback control is modeled by a sign function. To obtain resilience performance of the power system under low inertia and hybrid attacks, resilient distributed scheme combining load-frequency control (LFC) with virtual inertia control (VIC) is proposed. Then, resilient frequency control problem of the studied power system is converted to H8 control of the switched nonlinear system. By employing the Lyapunov stability theory and switched system method, the resilient conditions are given by the lower bound of the average dwell time of each residual topology and the upper bound of deception attacks. Furthermore, a linear matrix inequality (LMI) technique is used to design the distributed resilient control gains of the LFC-VIC scheme. Finally, a simulation of four-area power systems is carried out to verify the validness of our theory.
? Researchers have also investigated resilience to communication flaws (e.g., packet loss, delay, and link failure) in the cyber layer of microgrids.
? However, the scalability of the P2P configuration is largely constrained by the existing communication infrastructure.
? A hierarchical structure that is locally distributed and globally centralized is viewed as one of the most promising trends in this area.
? Other researchers have leveraged distributed peer-to-peer (P2P) communication protocols and noise-resilient state-observer techniques to achieve resilience against communication channel noise and disturbances.
? The proposed distributed algorithms rely on the communication network for exchanging information.
? Thus, failures in the communication network, caused by disconnection of communication links, cyber-attacks, delay and other channel imperfections can jeopardize the performance of the algorithms and lead to system-level reliability problems.
? This problem has not received much attention in smart power grids. Thus, this two-part paper aims to address this open problem for the distributed frequency regulation (DFR) algorithm.
? The DFR problem is formulated based on the quasi-steady state model of the prosumer power system.
• To improve the robustness of DSE under the presence of outliers, least-absolute-value (LAV) and generalized maximumlikelihood (GML) methods have been proposed.
• Regarding the host-based approach, an embedded intrusion detection method was proposed in for intelligent electronic devices (IED) in substations that monitor all incoming Generic Object-Oriented Substation Events (GOOSE) and Sampled Values (SV) messages to enhance cybersecurity.
• A hierarchical IDS framework was proposed in [84], which employed an SVM and an artificial immune system to analyze network traffic at every layer of a smart grid.
? The resilient DFR can provide many advantages for today’s power grids. For instance, it can complement today’s AGC systems, minimize system-wide control effort for regulating frequency, and require less sensing and communication compared with today’s centralized communication/control systems.
? The loads have a frequency-dependent component that will have an automatic beneficial effect without a controller.
? This frequency-dependent component can be modeled as damping coefficient.
? Under distributed frequency regulation, prosumers detect a frequency deviation and act accordingly to their physical characteristics such as droop and damping, effectively solving a distributed cost minimization.
|