Prediction accuracy of surrogate models as chaos indicator for nonlinear beam dynamics

Surrogate models have been widely used in studying nonlinear dynamical systems, including modern particle accelerators. A different perspective of viewing beam dynamical system from the prediction discrepancy of surrogate models is presented. The prediction accuracy was observed worse when particle motion is more chaotic. Therefore, such property can be used as a chaos indicator in beam dynamical system optimization. This new method was demonstrated with the National Synchrotron Light Source II storage ring.