Learning control of quantum systems using frequency-domain optimization algorithms

TL;DR

This paper explores frequency-domain optimization algorithms for quantum control, demonstrating their effectiveness in both known and unknown system models through atomic and molecular experiments.

Contribution

It introduces a frequency-domain gradient-based method for known quantum systems and a differential evolution approach for unknown systems, advancing quantum control techniques.

Findings

Successful population transfer in Rubidium atoms

Effective molecular fragmentation control in experiments

Demonstrated robustness and adaptability of algorithms

Abstract

We investigate two classes of quantum control problems by using frequency-domain optimization algorithms in the context of ultrafast laser control of quantum systems. In the first class, the system model is known and a frequency-domain gradient-based optimization algorithm is applied to searching for an optimal control field to selectively and robustly manipulate the population transfer in atomic Rubidium. The other class of quantum control problems involves an experimental system with an unknown model. In the case, we introduce a differential evolution algorithm with a mixed strategy to search for optimal control fields and demonstrate the capability in an ultrafast laser control experiment for the fragmentation of Pr(hfac)$_3$ molecules.