Berlin 2001 – scientific programme

Parts | Days | Selection | Search | Downloads | Help

Q: Quantenoptik

Q 23: Poster: Application of Short Pulses

Q 23.6: Poster

Thursday, April 5, 2001, 12:30–15:00, AT2

Evolutionary algorithms and their application to optimal control studies — •D. Zeidler, S. Frey, K.-L. Kompa, and M. Motzkus — Max-Planck-Institut für Quantenoptik, Hans-Kopfermann-Str. 1, D-85748 Garching, Germany

Coherent control of a physical or chemical process can be achieved by using modulated femtosecond (fs) laser pulses. A self-learning loop, which connects a fs pulse shaper, an optimization algorithm and an experimental feedback signal can automatically steer the interaction between system and laser field. The dependability of such a loop is crucial for the significance of the optimization results, which assigns the optimization algorithm a key role within these learning loops. An evolutionary strategy for fs pulse shaping is presented in detail. The stability and effectiveness of the algorithm is investigated both in experiments and simulations with emphasis on the influence of dimensionality of the parameter space, strategy parameters, and noise. The steering parameters which control the optimization behavior of the algorithm are non-critical if chosen within correct limits, while optimization speed and final result strongly depend on the number of free parameters and/or an appropriately chosen parameterization. Different parameterizations of the electric field which incorporate physical (or chemical) a-priori knowledge are presented and their influence on the optimization result is studied.