Multiobjective optimization of the dynamic aperture using surrogate models based on artificial neural networks

Modern synchrotron light source storage rings, such as the Swiss Light Source upgrade (SLS 2.0), use multibend achromats in their arc segments to achieve unprecedented brilliance. This performance comes at cost of increased focusing requirements, which turn require stronger sextupole and higher-order multipole fields for compensation effects on particles with energy deviation lead a considerable decrease dynamic aperture and/or acceptance. In this paper, increase these two quantities, multiobjective genetic algorithm (MOGA) is combined modified version well-known tracking code tracy. As first approach, massively parallel implementation MOGA used. Compared manually obtained solution approach yields very good results. However, it requires long computation time. second surrogate model based artificial neural networks used optimization. improves time, but quality results deteriorates beyond that solution. third retrained during ensures comparable one while also providing an order magnitude speedup. Finally, candidate solutions SLS 2.0 are shown further analyzed.