Author: Sapinski, M.
Paper Title Page
THPHA196 Automatized Optimization of Beam Lines Using Evolutionary Algorithms 1906
  • S. Appel, V. Chetvertkova, W. Geithner, F. Herfurth, U. Krause, S. Reimann, M. Sapinski, P. Schütt
    GSI, Darmstadt, Germany
  • D. Österle
    KIT, Karlsruhe, Germany
  Due to the massive parallel operation modes at the GSI accelerators, a lot of accelerator setup and re-adjustment has to be made during a beam time. This is typically done manually and is very time-consuming. With the FAIR project the complexity of the facility increases furthermore and for efficiency reasons it is recommended to establish a high level of automation. Modern Accelerator Control Systems allow a fast access to both, accelerator settings and beam diagnostics data. This provides the opportunity together with the fast-switching magnets in GSI-beamlines to implement evolutionary algorithms for automated adjustment. A lightweight python interface to CERN Front-End Software Architecture (FESA) gave the opportunity to try this novel idea, fast and easy at the CRYRING@ESR injector. Furthermore, the python interface facilitates the work flow significantly as the evolutionary algorithms python package DEAP could be used. DEAP has been applied already in external optimization studies with particle tracking codes*. The first results and gained experience of an automatized optimization at the CRYRING@ESR injector are presented here.
* S. Appel, O. Boine-Frankenheim, F. Petrov, Injection optimization in a heavy-ion synchrotron using genetic algorithms, Nucl. Instrum. Methods A, 852 (2017) pp. 73-79.
DOI • reference for this paper ※  
Export • reference for this paper using ※ BibTeX, ※ LaTeX, ※ Text/Word, ※ RIS, ※ EndNote (xml)