Researchers at Princeton and the Princeton Plasma Physics Laboratory have developed a machine learning method to control plasma edge bursts in fusion reactors, significantly enhancing performance without causing damage. This breakthrough allows for real-time optimization, reducing computation times from tens of seconds to milliseconds, and has been successfully tested at two different fusion facilities. The approach holds promise for future fusion devices, including ITER, and represents a significant advancement in the quest for sustainable fusion energy.
Researchers have successfully demonstrated the predictive control of fusion plasma using a digital twin control system that incorporates real-time observations and data assimilation. This new control approach, demonstrated at the Large Helical Device (LHD), allows for the accurate prediction and control of fusion plasma behavior, paving the way for advanced controls essential for the realization of fusion power generation. The system, developed by a research group led by Assistant Professor Yuya Morishita and Professor Sadayoshi Murakami of Kyoto University and other collaborators, is expected to become fundamental to fusion reactor control and has the potential to address other societal issues involving uncertain factors.
