Session K3

Keynote Speech: AI-Enabled Smart Grid Communications and Transactive Energy Systems

Conference
9:00 AM — 10:00 AM +08
Local
Oct 26 Wed, 9:00 PM — 10:00 PM EDT

Keynote Speech: AI-Enabled Smart Grid Communications and Transactive Energy Systems

Prof. Melike Erol-Kantarci, University of Ottawa, Canada

0
In the past decade, Information and Communication Technologies (ICT) have enabled the modernization of the power grid and have led to many advances in smart grid technologies. Smart grid communications facilitate a large number of grid operations, including advanced metering, fault monitoring, microgrid control, transactive energy systems and so on. In parallel to advances in smart grids, communication technologies have been continuously evolving to provide better service to mobile users and vertical industries. Recently, machine learning has showed promising performance improvements in communication networks as well as smart grid operations. In this talk, we introduce novel AI-based tools that will allow a P2P energy trading platform, consisting of microgrids, to become a part of the future transactive energy systems. The energy trading platform relies on robust smart grid communications. We will show our recent results on low-latency communications that use reinforcement learning to support communication needs of such energy trading platforms.

Session Chair and Room

Daisuke Mashima (Advanced Digital Sciences Center, Singapore) — Room LT2

Play Video Recording
Session K4

Keynote Speech: Machine Learning Based Security Solutions for Smart Grids: Challenges and Solutions

Conference
2:00 PM — 3:00 PM +08
Local
Oct 27 Thu, 2:00 AM — 3:00 AM EDT

Keynote Speech: Machine Learning Based Security Solutions for Smart Grids: Challenges and Solutions

Prof. Biplab Sikdar, National University of Singapore, Singapore

0
Smart grids take advantage of information and communication technologies to achieve energy efficiency, automation and reliability. Increasingly, smart grids are seeing a proliferation of dynamic new components and devices on the distribution edge of the grid. The integration of these components has led to new strategies for the planning and operational management of grids, particularly through two-way communications and power flow between the grid and consumers. However, these bidirectional communications and the convergence of the information technology and operational technology (IT/OT) networks introduce several security and privacy threats to the grid and the consumers. While machine learning based techniques have the potential to secure smart grids against various cyber threats, they are vulnerable to various attacks that can not only jeopardize the applications and their users, but also serve to expand the overall threat landscape. This talk will start by with examples of security vulnerabilities in machine learning applications in real-world smart grid environments and provide an overview of various types of attacks on machine learning algorithms. We will then provide a framework for security evaluation for machine learning in smart grid applications and present security solutions for such systems. The presentation will conclude with an overview of some promising approaches for future work in machine learning based security for smart grids.

Session Chair and Room

David Nicol (University of Illinois at Urbana-Champaign, United States) — Room LT2

Play Video Recording

Made with in Toronto · Privacy Policy · SmartGridComm 2020 · SmartGridComm 2021 · © 2022 Duetone Corp.