| Paper |
Title |
Page |
| WEPCH013 |
Electron Transport Line Optimization using Neural Networks and Genetic Algorithms
|
1948 |
| |
- D. Schirmer, T. Buening, P. Hartmann, D. Mueller
DELTA, Dortmund
|
|
| |
Methods of computational intelligence (CI) were investigated to support the optimization of the electron transfer efficiency from the booster synchrotron BoDo to the electron storage ring DELTA. Neural networks and genetic algorithms were analysed alternatively. At first both types of methods were trained on the basis of a theoretical model of the transport line. After the training various algorithms were used to improve the magnet settings of the real transport line elements with respect to the electron transfer efficiency. The results of different strategies are compared and prospects as well as limitations of CI-methods to the application of typical optimization problems in accelerator operation are discussed.
|
|