Summary

  • Researchers from the University of Texas have announced a new technique for calculating plasma behaviour in magnetic fields, which they claim outperforms existing methods.
  • The team used a combination of machine-learning techniques and traditional plasma simulations to develop a “data-driven symmetry theory”.
  • The new method could revolutionise the field of nuclear fusion, with the potential to create more efficient magnetic confinement devices that require lower energy inputs.
  • Existing methods, based on perturbative guiding centre models, have historically produced excellent results but the new, non-perturbative approach has the potential to improve optimisation of magnetic field configurations.
  • The findings were published in Physical Review Letters.

By Maya Posch

Original Article