Holistic Performance Benchmarking in Power Systems with Distributed Control under Disruptive Cyberattacks

Abstract

Distributed control architectures unlike centralized controllers in power systems do not have a single point of failure. To provide minimum failure and increased reliability and resiliency, utilizing a network of interconnected controllers, denoted as agents, for distributed control and management (power and energy) is necessary. Recently, the growing concern over malicious cyberattacks that can disrupt an entire system’s operation has further magnified distributed control architectures’ importance. Thus, it is imperative to evaluate and benchmark the performance of power systems with distributed control associated with algorithmic metric evaluation techniques. Such performance benchmarks enable better topological and methodological assessment of distributed control systems for engineers. This study proposes a novel approach in modeling performance metrics for power quality, reliability, and resiliency to develop a holistic normalized index, denoted as MetaMetric (MM), demonstrating a distributed control system’s effectiveness under cyberattack scenarios. Three types of cyberattacks are considered: 1) false data injection (FDI), 2) controller hijacking and 3) denial of service (DoS). The candidate system under study is a zonal MVDC ship power system with ten power management agents. The candidate system was designed for Naval applications, which have a higher susceptibility to malicious attacks than terrestrial systems.