Accelerating fusion science through learned plasma control

 

International Conference on Nuclear Physics






To solve the global energy crisis, researchers have long sought a source of clean, limitless energy. Nuclear fusion, the reaction that powers the stars of the universe, is one contender. By smashing and fusing hydrogen, a common element of seawater, the powerful process releases huge amounts of energy.









Here on earth, one way scientists have recreated these extreme conditions is by using a tokamak, a doughnut-shaped vacuum surrounded by magnetic coils, that is used to contain a plasma of hydrogen that is hotter than the core of the Sun. However, the plasmas in these machines are inherently unstable, making sustaining the process required for nuclear fusion a complex challenge. For example, a control system needs to coordinate the tokamak's many magnetic coils and adjust the voltage on them thousands of times per second to ensure the plasma never touches the walls of the vessel, which would result in heat loss and possibly damage. To help solve this problem and as part of DeepMind’s mission to advance science, we collaborated with the Swiss Plasma Center at EPFL to develop the first deep reinforcement learning (RL) system to autonomously discover how to control these coils and successfully contain the plasma in a tokamak, opening new avenues to advance nuclear fusion research.

In a paper published today in Nature, we describe how we can successfully control nuclear fusion plasma by building and running controllers on the Variable Configuration Tokamak (TCV) in Lausanne, Switzerland. Using a learning architecture that combines deep RL and a simulated environment, we produced controllers that can both keep the plasma steady and be used to accurately sculpt it into different shapes. This “plasma sculpting” shows the RL system has successfully controlled the superheated matter and - importantly - allows scientists to investigate how the plasma reacts under different conditions, improving our understanding of fusion reactors.
"In the last two years DeepMind has demonstrated AI’s potential to accelerate scientific progress and unlock entirely new avenues of research across biology, chemistry, mathematics and now physics."
Demis Hassabis, Co-founder and CEO, DeepMind


This work is another powerful example of how machine learning and expert communities can come together to tackle grand challenges and accelerate scientific discovery. Our team is hard at work applying this approach to fields as diverse as quantum chemistry, pure mathematics, material design, weather forecasting, and more, to solve fundamental problems and ensure AI benefits humanity.


#FusionAIAdvances #PlasmaControlRevolution #LearnedFusionScience #NextGenEnergy #AIinFusionResearch #FusionAccelerator #SmartPlasmaManipulation #AIempoweredFusion #PlasmaStabilityAI #FusionBreakthroughs



Comments

Post a Comment

Popular posts from this blog

Nuclear power: The safer and cheaper alternative to fossil fuels

Image
As countries across the globe struggle to balance their desire for a reliable energy supply with their contribution to climate change, nuclear power presents a solution. The global reliance on fossil fuels is a function of our energy needs, and although there have been numerous initiatives implemented in order to reduce demand, such as levying large taxes on air transport, appropriate sources of supply still need to be defined. Fossil fuels such as coal, oil and gas have the benefit of being capable of servicing our energy needs. However, they are a chief contributor to climate change, and as countries attempt to meet the Paris Agreement of keeping global temperature below 2o above pre-industrial levels, non-carbon energy sources are essential. Positives and negatives So, what are the pros and cons of nuclear power as a safer, cheaper alternative to fossil fuels? Well, where renewable sources such as wind or solar are clean, they are also intermittent, require an uneven geographical
Read more

Physicists Detect Subatomic Neutrinos Made by a Particle Collider for the First Time

Image
  International Conference on Nuclear Physics and Particle Physics Research Excellence In a scientific first, a team led by physicists at the University of California, Irvine (UCI) has detected neutrinos created by a particle collider. The discovery promises to deepen scientists’ understanding of the subatomic particles, which were first spotted in 1956 and play a key role in the process that makes stars burn. The work could also shed light on cosmic neutrinos that travel large distances and collide with the Earth, providing a window on distant parts of the universe. It’s the latest result from the Forward Search Experiment, or FASER, a particle detector designed and built by an international group of physicists and installed at CERN, the European Council for Nuclear Research in Geneva, Switzerland. There, FASER detects particles produced by CERN’s Large Hadron Collider. “We’ve discovered neutrinos from a brand-new source – particle colliders – where you have two beams of particles sma
Read more

Physics Breakthrough: Scientists Discover Rydberg Moiré Excitons

Image
International Conference on Nuclear Physics The Rydberg state is prevalent across various physical mediums such as atoms, molecules, and solid materials. Rydberg excitons, which are highly energized, Coulomb-bound electron-hole pair states, were initially identified in the 1950s within the semiconductor material, Cu2O. In a study published in Science, Dr. Xu Yang and his colleagues from the Institute of Physics (IOP) of the Chinese Academy of Sciences (CAS), in collaboration with researchers led by Dr. Yuan Shengjun of Wuhan University, have reported observing Rydberg moiré excitons, which are moiré-trapped Rydberg excitons in the monolayer semiconductor WSe2 adjacent to small-angle twisted bilayer graphene (TBG). The solid-state nature of Rydberg excitons, combined with their large dipole moments, strong mutual interactions, and greatly enhanced interactions with the surroundings, holds promise for a wide range of applications in sensing, quantum optics, and quantum simulation. Howeve
Read more