Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

arXiv:quant-ph/0401018 · Journal of Physics B Atomic Molecular and Optical Physics

Extracting quantum dynamics from genetic learning algorithms through principal control analysis

J L White, B J Pearson, P. H. Bucksbaum

Published Dec 7, 2004• 57 citations
arXivHTMLPDF

Abstract

Genetic learning algorithms are widely used to control ultrafast optical pulse shapes for photo-induced quantum control of atoms and molecules. An unresolved issue is how to use the solutions found by these algorithms to learn about the system's quantum dynamics. We propose a simple method based on covariance analysis of the control space, which can reveal the degrees of freedom in the effective control Hamiltonian. We have applied this technique to stimulated Raman scattering in liquid methanol. A simple model of two-mode stimulated Raman scattering is consistent with the results.

DOI
10.1088/0953-4075/37/24/l02

Related papers

Extracting quantum dynamics from genetic learning algorithms through principal component analysis

J. L. White, Brett J. Pearson, P. H. Bucksbaum

Genetic learning algorithms are widely used to control ultrafast optical pulse shapes for photo-induced quantum control of atoms and molecules. An outstanding issue is how to use the solutions found by these algorithms to learn about the system's quantum dynamics. We propose a simple method based on principal component analysis of the control space, which can reveal the degrees of freedom responsible for control, and aid in the construction of an effective Hamiltonian for the dynamics.

•Jan 6, 2004•0 citations

Learning from molecules: Coherent feedback control in liquid phase.

Brett J. Pearson

Coherent control of light and matter is a major tool in current atomic physics and physical chemistry research. Technological advances in ultrafast laser pulse shaping allow control over the dynamics of molecular systems on sub-picosecond timescales. This thesis demonstrates feedback control over a variety of quantum systems and describes an adaptive genetic algorithm used in the experiments. In the gas phase, we examine control of multi-photon, dissociative ionization in diatomic sodium. The learning algorithm achieves control over the dissociative fraction primarily by adjusting the peak intensity of the laser field. We also investigate control of multi-mode stimulated Raman scattering in liquid phase methanol. Analysis of the results leads to identification of quasi-impulsive Raman coupling as a possible control mechanism. Further experiments, combined with numerical simulations of the interaction, demonstrate that an impulsive coupling of the two modes by the pump pulse can control the Raman gain. We extend the Raman control experiments to ethanol and isopropanol. Finally we apply feedback control techniques in an initial investigation of a liquid-phase chemical reaction. Specifically, we examine the ring-opening conversion of cyclohexadiene into hexatriene. The reaction dynamics occur on an excited state potential energy surface. Through control of the excitation pulse shape, we seek to guide the wavepacket evolution on the excited state to exert control over the reaction branching ratio.

•Jan 1, 2004•0 citations

Using feedback for coherent control of quantum systems

T. C. Weinacht, P. H. Bucksbaum

A longstanding goal in chemical physics has been the control of atoms and molecules using coherent light fields. This paper provides a brief overview of the field and discusses experiments that use a programmable pulse shaper to control the quantum state of electronic wavepackets in Rydberg atoms and electronic and nuclear dynamics in molecular liquids. The shape of Rydberg wavepackets was controlled by using tailored ultrafast pulses to excite a beam of caesium atoms. The quantum state of these atoms was measured using holographic techniques borrowed from optics. The experiments with molecular liquids involved the construction of an automated learning machine. A genetic algorithm directed the choice of shaped pulses which interacted with the molecular system inside a learning control loop. Analysis of successful pulse shapes that were found by using the genetic algorithm yield insight into the systems being controlled.

Journal of Optics B Quantum and Semiclassical Optics•Apr 3, 2002•60 citations

Using Feedback for Coherent Control of Quantum Systems

Thomas C. Weinacht

The control of atoms and molecules using coherent light fields has been a long-standing goal in chemical physics. This thesis discusses experiments which use a programmable pulse shaper to control the quantum state of electronic wavepackets in Rydberg atoms and electronic and nuclear dynamics in molecular liquids. The shape of Rydberg wavepackets was controlled by using tailored ultrafast pulses to excite a beam of cesium atoms. The quantum state of these atoms was measured using holographic techniques borrowed from optics. The experiments with molecular liquids involved the construction of an automated learning machine. A Genetic Algorithm directed the choice of shaped pulses which interacted with the molecular system inside a learning control loop. Analysis of successful pulse shapes that were found by using the GA yield insight into the systems being controlled.

APS Division of Atomic, Molecular and Optical Physics Meeting Abstracts•May 1, 2001•0 citations

Accelerated optimization and automated discovery with covariance matrix adaptation for experimental quantum control

Jonathan Roslund, Ofer M. Shir, Thomas Bäck +1 more

Optimization of quantum systems by closed-loop adaptive pulse shaping offers a rich domain for the development and application of specialized evolutionary algorithms. Derandomized evolution strategies (DESs) are presented here as a robust class of optimizers for experimental quantum control. The combination of stochastic and quasi-local search embodied by these algorithms is especially amenable to the inherent topology of quantum control landscapes. Implementation of DES in the laboratory results in efficiency gains of up to $\ensuremath{\sim}9$ times that of the standard genetic algorithm, and thus is a promising tool for optimization of unstable or fragile systems. The statistical learning upon which these algorithms are predicated also provide the means for obtaining a control problem's Hessian matrix with no additional experimental overhead. The forced optimal covariance adaptive learning (FOCAL) method is introduced to enable retrieval of the Hessian matrix, which can reveal information about the landscape's local structure and dynamic mechanism. Exploitation of such algorithms in quantum control experiments should enhance their efficiency and provide additional fundamental insights.

Physical Review A•Oct 27, 2009•43 citations

Rovibrational wave-packet manipulation using shaped midinfrared femtosecond pulses toward quantum computation: Optimization of pulse shape by a genetic algorithm

Masaaki Tsubouchi, Takamasa Momose

We have numerically examined optimization of the experimentally achievable shapes of mid-infrared femtosecond laser pulses to construct elementary quantum gates using molecular rovibrational states of a single molecule. Instead of optimizing the electric field by optimal control theory, the method generally used in theoretical studies of coherent control, the transmittance and the phase shift of conventional pulse shapers were optimized in the frequency domain by a genetic algorithm. The target molecular system we examined was the rotation-vibration states of CO in the $X({^{1}\ensuremath{\Sigma}}^{+})$ ground electronic state. Although the existence of the rotational degrees of freedom makes the quantum gate operation complicated, high fidelity of over 0.95 was achieved for one-rovibrational-qubit systems, which indicates that a molecular quantum computer may be feasible for at least one-qubit calculations on a diatomic molecule. On the other hand, two-qubit calculations are more difficult to achieve in a single CO molecule using rotational and vibrational degrees of freedom as different qubits. The importance of the control of the rotational in addition to the vibrational wave packet in order to realize a molecular quantum computer is discussed.

Physical Review A•May 20, 2008•57 citations
Genetic learning algorithms are widely used to control ultrafast optical pulse shapes for photo-induced quantum control of atoms and molecules. An unresolved issue is how to use the solutions found by these algorithms to learn about the system's quantum dynamics. We propose a simple method based on covariance analysis of the control space, which can reveal the degrees of freedom in the effective control Hamiltonian. We have applied this technique to stimulated Raman scattering in liquid methanol. A simple model of two-mode stimulated Raman scattering is consistent with the results.
Coherent control of light and matter is a major tool in current atomic physics and physical chemistry research. Technological advances in ultrafast laser pulse shaping allow control over the dynamics of molecular systems on sub-picosecond timescales. This thesis demonstrates feedback control over a variety of quantum systems and describes an adaptive genetic algorithm used in the experiments. In the gas phase, we examine control of multi-photon, dissociative ionization in diatomic sodium. The learning algorithm achieves control over the dissociative fraction primarily by adjusting the peak intensity of the laser field. We also investigate control of multi-mode stimulated Raman scattering in liquid phase methanol. Analysis of the results leads to identification of quasi-impulsive Raman coupling as a possible control mechanism. Further experiments, combined with numerical simulations of the interaction, demonstrate that an impulsive coupling of the two modes by the pump pulse can control the Raman gain. We extend the Raman control experiments to ethanol and isopropanol. Finally we apply feedback control techniques in an initial investigation of a liquid-phase chemical reaction. Specifically, we examine the ring-opening conversion of cyclohexadiene into hexatriene. The reaction dynamics occur on an excited state potential energy surface. Through control of the excitation pulse shape, we seek to guide the wavepacket evolution on the excited state to exert control over the reaction branching ratio.
A longstanding goal in chemical physics has been the control of atoms and molecules using coherent light fields. This paper provides a brief overview of the field and discusses experiments that use a programmable pulse shaper to control the quantum state of electronic wavepackets in Rydberg atoms and electronic and nuclear dynamics in molecular liquids. The shape of Rydberg wavepackets was controlled by using tailored ultrafast pulses to excite a beam of caesium atoms. The quantum state of these atoms was measured using holographic techniques borrowed from optics. The experiments with molecular liquids involved the construction of an automated learning machine. A genetic algorithm directed the choice of shaped pulses which interacted with the molecular system inside a learning control loop. Analysis of successful pulse shapes that were found by using the genetic algorithm yield insight into the systems being controlled.
The control of atoms and molecules using coherent light fields has been a long-standing goal in chemical physics. This thesis discusses experiments which use a programmable pulse shaper to control the quantum state of electronic wavepackets in Rydberg atoms and electronic and nuclear dynamics in molecular liquids. The shape of Rydberg wavepackets was controlled by using tailored ultrafast pulses to excite a beam of cesium atoms. The quantum state of these atoms was measured using holographic techniques borrowed from optics. The experiments with molecular liquids involved the construction of an automated learning machine. A Genetic Algorithm directed the choice of shaped pulses which interacted with the molecular system inside a learning control loop. Analysis of successful pulse shapes that were found by using the GA yield insight into the systems being controlled.
Optimization of quantum systems by closed-loop adaptive pulse shaping offers a rich domain for the development and application of specialized evolutionary algorithms. Derandomized evolution strategies (DESs) are presented here as a robust class of optimizers for experimental quantum control. The combination of stochastic and quasi-local search embodied by these algorithms is especially amenable to the inherent topology of quantum control landscapes. Implementation of DES in the laboratory results in efficiency gains of up to $\ensuremath{\sim}9$ times that of the standard genetic algorithm, and thus is a promising tool for optimization of unstable or fragile systems. The statistical learning upon which these algorithms are predicated also provide the means for obtaining a control problem's Hessian matrix with no additional experimental overhead. The forced optimal covariance adaptive learning (FOCAL) method is introduced to enable retrieval of the Hessian matrix, which can reveal information about the landscape's local structure and dynamic mechanism. Exploitation of such algorithms in quantum control experiments should enhance their efficiency and provide additional fundamental insights.
We have numerically examined optimization of the experimentally achievable shapes of mid-infrared femtosecond laser pulses to construct elementary quantum gates using molecular rovibrational states of a single molecule. Instead of optimizing the electric field by optimal control theory, the method generally used in theoretical studies of coherent control, the transmittance and the phase shift of conventional pulse shapers were optimized in the frequency domain by a genetic algorithm. The target molecular system we examined was the rotation-vibration states of CO in the $X({^{1}\ensuremath{\Sigma}}^{+})$ ground electronic state. Although the existence of the rotational degrees of freedom makes the quantum gate operation complicated, high fidelity of over 0.95 was achieved for one-rovibrational-qubit systems, which indicates that a molecular quantum computer may be feasible for at least one-qubit calculations on a diatomic molecule. On the other hand, two-qubit calculations are more difficult to achieve in a single CO molecule using rotational and vibrational degrees of freedom as different qubits. The importance of the control of the rotational in addition to the vibrational wave packet in order to realize a molecular quantum computer is discussed.