Computer scientists and experts in energy technology at Case Western Research University have come together to promote artificial intelligence or diagnostic power AI to make solar power plants more efficient.
Solar energy uses energy coming from the sun collected by photovoltaic or PV modules to develop clean and renewable energy.
Researchers said making more solar energy plants more efficient will benefit the industry and, indeed, consumers.
According to the Solar Durability and Lifetime Extension Research Center and Kyocera Professor of Ceramic, Department of Materials Science and Engineering at the director of the Roger French School of Case Engineering, supply is currently the “cheapest form of electricity in the world.” However, the efficiency of power plants is being evaluated at the same time, and that is not just “tractable, especially for a fast-growing industry.”
This study, funded by three years or the equivalent of a $ 750,000 grant from the U.S. Department of Energy, as part of the DOE-announced $ 130 million solar technologies comprehensive effort, includes $ 7.3 million, primarily for machine learning. campaigns, as well as other artificial intelligence for solar applications.
Associate professor of French research and Materials Science and Engineering Laura Bruckman are co-authors of the study.
CENTRAL RESPONSE: Exposed solar cells will soon enable on-site energy generation
(Photo: Sean Gallup / Getty Images)
The research team said it would work on collecting and evaluating data this year, and then start offering a ‘pre-trained computer model’ to individual solar energy companies and power plants. to explore how to develop their own system.
Machine learning for more efficient solar energy
Essentially, this study led by Western Reserve aims to use technology, particularly computers, better to evaluate data from large volumes of PV systems to help quantify short-term performance. long-term and long-term.
These machine learning techniques will be used to resolve data quality issues affecting the individual plants. For that to happen, the study’s authors say they would use a “spatiotemporal graph neural network framework.”
The spatiotemporal method involves determining how plants perform differently in space and time and building a framework to make the specific PV plants in that group and systems in time to improve future.
Mehmet Koyuturk, professor of Computer Science at Andrew R. Jennings and one of the team members, said there is no robot to visit all the “photovoltaic plants to look at information and confirm patterns” of similarity between behaviors, all of the collected data used instead, to function “as it were.”
Pre-training computer model for solar energy companies
According to Bruckman, different companies have information about their own technology in a specific area of their country. However, so far, “we have not had the opportunity to be able to collect and evaluate” all data from a wide range of fields and companies.
Finally, the team, researchers, and Yinghui Wu, one of the team members, assistant professor of the Department of Computer and Data Sciences, said the study would help the solar industry and ultimately energy consumers, but researchers AI, too.
Every time a new system for understanding new data from specific fields is understood, explained Wu, who is also a co – researcher on research funded by the National Science Foundation, it helps to further understand “our own science. “
French explained that his group would work on collecting and evaluating data this year, and then start offering a “pre-training computer model” to individual solar energy companies and power plants to explore how to develop their own system. ‘
RELATED SUBJECT: Dissolving baths promise better perovskite crystals and better solar energy
Check out more news and information about Solar Energy on Science Times.