http://www.energy-ai.org/
Plenary speaker
Digital Twins as a key enabling technology for AI Powered Cyber-physical Power/Energy Systems
AI powered cyber-physical power/energy systems hold a great promise for the future. Who would not love to have intelligent, learning, adaptive system capable to exchange data and learn from its environment? Moreover, who would not love to have such systems at all levels of control hierarchy: from the levels that are the closest to the “metal” to the levels that are farthest away from the “metal”?
From the point of view of AI application in Cyber-physical Power/Energy systems a completely safe and data rich environment for AI to learn-by-doing is a challenge that can only be solved by judicious application of high-performance Digital Twin technology. For the purpose of this keynote we will treat the control levels that are very close to the “metal” as mature and part of the digital twin that is ready to support the development of higher level AI functions like the already demonstrated examples Deepmind’s AlphaStar and OpenAI Five that excel in most complex, real time, strategy games like Star Craft 2 and Dota 2.
Clearly, to unleash the power of machine learning on Power/Energy Systems a good “sandbox” is a must. So, what defines a good sandbox? A good “sandbox” has excellent physical models, that execute in real time which enables them to interface directly to the control layers that are closest to the “metal”. Such a system can then easily be interfaced to the AI algorithm and “left” to learn how to operate it until “perfection”. Finally, an AI controller verified on a digital twin can be “readily” deployed in the real e.g. grid when there are no more exceptional cases to be verified.
Biography
Nikola is the CEO and co-founder of Typhoon-HIL, and was the leader of the team that developed both the theoretical algorithms and experimental validation of the world’s first 1μs ultra-low latency Hardware-in-the Loop (HIL) real-time emulator platform for power electronics.
Prior to founding Typhoon HIL he was with the ABB Drives Development Department, Turgi, Switzerland. Previously he was with the ABB Research Center, Baden-Dättwil, Switzerland.
He holds a Ph.D. degree from Virginia Polytechnic Institute and State University, Blacksburg in 1995, an M.S. degree in mechanical engineering from the Vanderbilt University, Nashville, TN, and a B.S. degree in electrical engineering from the University of Novi Sad, Serbia.
Energy and AI 是天津大学官方期刊，其目标是成为能源和人工智能交叉领域前沿研究的权威期刊。该期刊主要刊登研究(Research)、短讯(Short communication)、展望(Perspective)、综述(Review)等类型的文章。欢迎广大学者踊跃投稿。我们的网址是：www.journals.elsevier.com/energy-and-ai
