Learning-Based Quantum Robust Control: Algorithm, Applications, and Experiments.

Area of Science:

• Quantum Information Technology
• Atomic Physics
• Molecular Chemistry

Background:

• Robust control is crucial for advancing quantum information technology, molecular chemistry, and atomic physics.
• Existing methods face challenges in optimizing complex quantum systems with uncertainties.

Purpose of the Study:

• To introduce an improved differential evolution algorithm, msMS_DE, for robust quantum control.
• To apply msMS_DE to challenging quantum control problems, including open quantum ensembles and networked quantum systems.
• To validate the algorithm's efficacy through experimental implementation in laser-based molecular control.

Main Methods:

• Developed a novel multiple-samples and mixed-strategy DE (msMS_DE) algorithm.
• Utilized multiple samples for fitness evaluation and a mixed strategy for mutation operations.
• Applied msMS_DE to numerical simulations of quantum ensembles and networks, and experimental femtosecond laser control.

Main Results:

• msMS_DE demonstrated superior performance in numerical simulations for robust control of quantum ensembles and networks.
• Experimental results confirmed msMS_DE's effectiveness in optimizing two-photon absorption and controlling molecular fragmentation (CH2BrI) using femtosecond laser pulses.
• The algorithm successfully identified effective laser pulse parameters for complex quantum tasks.

Conclusions:

• The msMS_DE algorithm offers a powerful and effective approach for robust quantum control.
• This method significantly advances the practical application of quantum control in diverse scientific fields.
• The successful experimental validation highlights the algorithm's potential for real-world quantum technology development.