Self-organising smart grid architectures for cyber-security

Abstract

Current conventional power systems consist of large-scale centralised generation and unidirectional power flow from generation to demand. This vision for power system design is being challenged by the need to satisfy the energy trilemma, as the system is required to be sustainable, available and secure. Emerging technologies are restructuring the power system; the addition of distributed generation, energy storage and active participation of customers are changing the roles and requirements of the distribution network. Increased controllability and monitoring requirements combined with an increase in controllable technologies has played a pivotal role in the transition towards smart grids. The smart grid concept features a large amount of sensing and monitoring equipment sharing large volumes of information. This increased reliance on the ICT infrastructure, raises the importance of cyber-security due to the number of vulnerabilities which can be exploited by an adversary. The aim of this research was to address the issue of cyber-security within a smart grid context through the application of self-organising communication architectures. The work examined the relevance and potential for self-organisation when performing voltage control in the presence of a denial of service attack event. The devised self-organising architecture used techniques adapted from a range of research domains including underwater sensor networks, wireless communications and smart-vehicle tracking applications. These components were redesigned for a smart grid application and supported by the development of a fuzzy based decision making engine. A multi-agent system was selected as the source platform for delivering the self-organising architecture The application of self-organisation for cyber-security within a smart grid context is a novel research area and one which presents a wide range of potential benefits for a future power system. The results indicated that the developed self-organising architecture was able to avoid control deterioration during an attack event involving up to 24% of the customer population. Furthermore, the system also reduces the communication load on the agents involved in the architecture and demonstrated wider reaching benefits beyond performing voltage control.