Configurable UAL-based modeling engine for online accelerator studies
N. Malitsky, Todd J. Satogata, Richard M. Talman · 2004
The diversity of online accelerator modeling tasks poses a considerable challenge for accelerator physicists and software developers. Compromises between performance and completeness are always required. For example, modeling of chromatic effects requires second-order calculations, while speed favors first-order matrix multiplication. This paper presents our solution: a configurable computational modeling engine based on the UAL Accelerator Propagator Framework (APF). The choices of evolution algorithms are defined in an external Accelerator Propagator Description Format (APDF) file, permitting a flexible mechanism for employing different approaches within the context of machine studies and operation.