Reviewed by Lexie CornerFeb 29 2024
In a world where electrification is increasingly sourced from variable sources, such as solar and wind power, researchers have reported the development of artificial intelligence algorithms designed to respond quickly when the network’s voltage balance is threatened.
Open-access AI algorithms have been developed to protect electrical grids from random fluctuations introduced by renewable energy and electric vehicles. Assistant Professor Qianwen Xu in her lab at the Department of Electric Power and Energy Systems. Image Credit: KTH Royal Institute of Technology
While electric vehicles and renewable energy may be better for the environment, their combination has the potential to upset power systems and cause a host of issues, from broken laptops to local blackouts. This is because sporadic fluctuations in supply and demand strain the network’s ability to keep voltage levels constant.
According to Qianwen Xu, a researcher at Stockholm’s KTH Royal Institute of Technology, an open-source AI solution was created to alleviate this pressure.
Wind power and solar radiation are not consistent from hour to hour, and demand for charging EVs is based on people’s personal needs and habits. So, you have a high level of stochastics and uncertainties. Their integration will lead to voltage fluctuations, deviations, and even voltage security violation challenges.
Qianwen Xu, Researcher, KTH Royal Institute of Technology
According to Xu, the new open-source deep reinforced learning (DRL) algorithms address this difficulty by providing intelligence to power converters located far inside the grid, where they safely optimize large-scale energy source coordination under sudden fluctuations without requiring real-time communication. The DRL offers a unique data synchronization technique for data-driven algorithms to address communication delays.
Fast and Cost-Effective
Centralized control is not cost-efficient or fast under continuous fluctuations of renewable energy and electric vehicles. Our aim is an AI-based self control for each distributed energy source, which are interfaced by power converters.
Qianwen Xu, Researcher, KTH Royal Institute of Technology
The researchers showcased this on a real-world smart microgrid hardware platform at KTH. The open-source software package has been published on GitHub. The study was published in the journal IEEE Transactions on Sustainable Energy.
The decentralized management strategy of the solution would keep voltage levels within predetermined bounds. According to Xu, voltage fluctuations that exceed that margin run the risk of negatively affecting the grid's overall stability and the electrical equipment's functionality.
Xu says that voltage deviations can trigger the inefficient operation of electrical devices, shorten their lifespan, and, in severe instances, inflict damage on the grid infrastructure. Voltage security breaches can also escalate to blackouts or necessitate emergency measures like load shedding or the deployment of reserve generators to uphold grid stability.
Xu says, “Our purpose is to improve control strategies for power converters by making them more adaptive and intelligent in order to stabilize complex and changing power grids.”
This work is part of Digital Futures, a research center at KTH that collaborates with scientists from Stockholm University, the University of California, Berkeley, and other institutions to explore and develop digital technologies.
Journal Reference:
Zhang, M., et al. (2024) Data Driven Decentralized Control of Inverter based Renewable Energy Sources using Safe Guaranteed Multi-Agent Deep Reinforcement Learning. IEEE Transactions on Sustainable Energy. doi.org/10.1109/TSTE.2023.3341632.