An anonymous reader cites a report from Princeton Engineering. The uncontrolled, superheated plasma that drives the fusion reaction can quickly lose its stability and escape the powerful magnetic fields that trap it inside the doughnut-shaped fusion reactor. Such escape often marks the end of a reaction, and is a central challenge in developing fusion as a pollution-free and virtually limitless source of energy. But a Princeton-led team of engineers, physicists, and data scientists from Princeton University and the Princeton Plasma Physics Laboratory (PPPL) is harnessing the power of artificial intelligence to predict the occurrence of specific plasma problems. Avoided. in real time.
In an experiment at the DIII-D National Fusion Facility in San Diego, researchers found that a model trained only on historical experimental data detected potential plasma instability, known as tearing mode instability, by up to 300 milliseconds in advance. It has been demonstrated that it is possible to predict This leaves just enough time for a human to blink slowly, but the AI controller changes certain operating parameters to create cracks in the plasma's magnetic field lines, disrupting the plasma's equilibrium and aperture. That was enough time to avoid putting it away. The door to escape after the reaction ends.
“By learning from past experiments, rather than incorporating information from physics-based models, AI has the potential to develop final control policies that support stable, high-power plasma conditions in real-time in real reactors. “Yes,” said study leader Egemen Colemen. , an associate professor in Mechanical and Aerospace Engineering and the Andlinger Center for Energy and the Environment, and a staff research physicist at PPPL. This work opens the door for more dynamic control of fusion reactions than current approaches and artificially eliminates the wide range of plasma instabilities that have long been an obstacle to achieving sustained fusion reactions. It provides a foundation for solving using intelligence. The research team published their findings in his journal Nature.