Modeling Accelerator Magnets’ History Using ML

A magnet’s past will allow for the customization of particle beams precisely in the future. As accelerators stretch for higher performance, understanding subtle effects, such as those introduced by magnetic history, is becoming more critical.

A team of researchers at the Department of Energy’s SLAC National Accelerator Laboratory and other institutions developed a powerful mathematical technique using machine learning to model a magnet’s previous states and make predictions about future states. The approach eliminates the need to reset the magnets and immediately improves accelerator performance.

The model looks at an important property of magnets known as hysteresis—residual or leftover magnetism. The new model removes the need to reset magnets as often and can enable both machine operators and automated tuning algorithms to quickly see their present state, making what was once invisible visible.

The hysteresis model could also help smaller accelerator facilities, which might not have as many researchers and engineers to reset magnets and run higher-precision experiments. The team hopes to implement the method across a full set of magnets at an accelerator facility and demonstrate an improvement in predictive accuracy on an operational accelerator.