This thesis presents the implementation of a control system for an ultra-cold atom experiment. This system allows for the acceleration of data acquisition and the use of machine learning algorithms to optimize the operation of the experiment. The characteristics of the new system are presented, as well as the software suite allowing the control and analysis of the data. The different algorithms used are described, as well as the first optimization results obtained.In the experiment, a Bose-Einstein condensate of rubidium atoms is confined to the surface of an ellipsoid of revolution. This original trapping geometry comes from the dressing of the atoms by a radiofrequency field in a magnetic quadrupole trap.In this geometry, it is possible to compensate for the effect of gravity, and I present the experimental results obtained. We observe the spontaneous formation of a condensed ring near the equator of the bubble, and we interpret this behavior as an effect of the zero point energy of the transverse confinement.A more elaborate model requiring the introduction of terms beyond the rotating wave approximation is compared to the experimental data.
This thesis presents the implementation of a control system for an ultra-cold atom experiment. This system allows for the acceleration of data acquisition and the use of machine learning algorithms to optimize the operation of the experiment. The characteristics of the new system are presented, as well as the software suite allowing the control and analysis of the data. The different algorithms used are described, as well as the first optimization results obtained.In the experiment, a Bose-Einstein condensate of rubidium atoms is confined to the surface of an ellipsoid of revolution. This original trapping geometry comes from the dressing of the atoms by a radiofrequency field in a magnetic quadrupole trap.In this geometry, it is possible to compensate for the effect of gravity, and I present the experimental results obtained. We observe the spontaneous formation of a condensed ring near the equator of the bubble, and we interpret this behavior as an effect of the zero point energy of the transverse confinement.A more elaborate model requiring the introduction of terms beyond the rotating wave approximation is compared to the experimental data.
Control and optimization of a Bose-Einstein condensation experiment Dans cette thèse est présentée l’implémentation d’un système de contrôle d’une expérience d’atomes ultra-froids. Celui-ci permet d’accélérer l’acquisition de données et d’utiliser des algorithmes d’apprentissage automatique pour optimiser le fonctionnement de l’expérience. Les caractéristiques du nouveau système sont présentées ainsi que la suite logicielle permet-tant le contrôle et l’analyse des données. Les différents algorithmes utilisés sont décrits ainsi que les premiers résultats d’optimisation obtenus. Dans l’expérience, un condensat de Bose-Einstein de gaz d’atomes de rubidium est confiné à la surface d’un ellipsoïde de révolution. Cette géométrie originale de piégeage provient de l’habillage des atomes par un champ radiofréquence dans un piège quadrupolaire magnétique. Dans cette géométrie il est possible de compenser l’effet de la gravité, et je présente les résultats expérimentaux obtenus. Nous observons la formation spontanée d’un anneau condensé proche de l’équateur de la bulle, et nous interprétons ce comportement par un effet de l’énergie de point zéro du confinement transverse. Un modèle plus élaboré nécessitant l’introduction de termes au-delà de l’approximation de l’onde tournante est comparé aux données expérimentales.
In this thesis we use optimal control theory to control a BoseEinstein condensate. We focus on an experiment conducted at the Vienna University of Technology where a Bose-Einstein condensate is produced and trapped on an atom chip, and two entangled matter waves are produced. Such a twin atom beam has many technical applications and is necessary for matter wave interferometry, for example. One critical step in this experiment is the excitation of the condensate, which is achieved by displacing the minimum of the confinement potential (“shaking”). In order to find a time dependent potential that excites the condensate, optimal control theory is necessary and able to produce satisfying results. However, theoretical predictions and experimental data do not match perfectly. This is mainly due to the impulse response filter, causing a difference between the calculated and the experimentally accessible control field. To solve this problem in this thesis, we incorporate the effect of this impulse response filter in the optimal control algorithm. The new system is then not local in time anymore and there are new boundary conditions to meet. These problems are overcome and the obtained control fields are much closer to the experiment than without considering the effects of the impulse response filter. It is even possible to find working control fields with a transition time of only two milliseconds or to transfer the condensate in a higher excited state.
Bose-Einstein condensates (BECs), a quantum state of matter formed when bosonic atoms are cooled close to absolute zero, have become the premier platform for investigating fundamental physics with atomic vapours. Experiments on Bose-Einstein condensates now achieve exquisite control over many aspects of the system, including interactions, trapping potential, and dynamics. This has precipitated a new wave of research into many-body quantum phenomena and, in particular, solitons. These structures are fundamental excitations of an interacting non-linear medium, of interest to a multitude of scientific disciplines from non-linear optics to financial markets. The highly controllable environment of BECs form an attractive playground for the study of solitons allowing the non-linearity to be dynamically tuned, facilitating deeper investigations into these structures. Consistently generating and analysing solitons in BEC experiments continues to be problematic. In particular, the non-linear dynamics of BECs, though required for the generation of solitons, produce particularly challenging control and optimization problems. These control problems must be solved before further investigations into the fundamental physics of soliton dynamics can be answered. This thesis makes three important advances in the control and measurement of BECs that will lead to better generation and observation of solitons. (1) a theoretical model for a control scheme capable of highly precise wavefunction engineering, (2) the experimental implementation of a machine learning algorithm for online optimisation, and (3) a continuous non-destructive imaging system capable of directly observing soliton dynamics in real-time. Together, these advances provide a suite of tools for manipulating and exploiting solitons in Bose-Einstein condensates. A novel technique was developed theoretically, offering control of the macroscopic wavefunction of a Bose-Einstein condensate with unprecedented spatial resolution and speed. The ability to control the atomic wavefunction at the fundamental length scale is key to the advancement of many quantum technologies such as quantum simulators. The magnetic resonance control scheme is demonstrated through simulation of a 87Rb condensate with the exemplar model generating a single dark soliton with corresponding π phase kink. The soliton represents a structure at the fundamental length scale of the system, and demonstrates the potential of the scheme for precision state engineering. The scheme is extended to generate higher-order soliton modes which are yet to be experimentally realised. A machine learning algorithm based on Gaussian processes was developed and implemented on the evaporative cooling stage of the production of a 87Rb Bose-Einstein condensate, successfully demonstrating fast optimisation to condensation. The Gaussian process develops a statistical model based on the data that enables the characterisation of the relationship between the experimental controls and resultant quality of the BEC. This relationship is often obfuscated through technical details of the apparatus, frustrating the use of theoretical models to design optimal evaporation ramps. These models often only consider ergodic dynamics with two-body s-wave interactions and no other loss rates with better ramps likely exploiting more complex interactions. The internal model generated from the Gaussian process utilised uncertainty in measured data, making the optimisation more robust to experimental noise than alternate methods. The algorithm is shown to produce high quality Bose-Einstein condensates in 10 times fewer experimental iterations than previously used online optimisation techniques. By exploiting information on the sensitivity of each control, the model can be used to aid experimental design. The convergence of the optimisation is further improved by eliminating a superfluous parameter identified by the model. The general usefulness of machine learning compared with bespoke optimisation algorithms has seen machine learning approach ubiquity. Finally, an experimentally straightforward technique for continuous non-destructive imaging of matter-wave solitons was developed and implemented, facilitating measurements of stochastic phenomena. The technique is readily practicable on any ultracold atom experiment with an existing absorption imaging system, simply requiring the probe laser be far-detuned from resonance. With a signal-to-noise of ∼ 33 at 1.25 GHz detuning, the technique is capable of producing 100 images with no observable heating or atom loss. Coupled with a fast optical phase locked loop, the technique can be used in conjunction with absorption imaging to generate a series of non-destructive images followed by a final high signal-to-noise absorption image solely through moving the laser on resonance for the final image. The high performance and utility of this imaging setup make it a powerful tool for ultra-cold atom experiments.
Machine learning is emerging as a technology that can enhance physics experiment execution and data analysis. Here, we apply machine learning to accelerate the production of a Bose-Einstein condensate (BEC) of 87Rb atoms by Bayesian optimization of up to 55 control parameters. This approach enables us to prepare BECs of $2.8\ifmmode\times\else\texttimes\fi{}{10}^{3}$ optically trapped 87Rb atoms from a room-temperature gas in 575 ms. The algorithm achieves the fast BEC preparation by applying highly efficient Raman cooling to near quantum degeneracy, followed by a brief final evaporation. We anticipate that many other physics experiments with complex nonlinear system dynamics can be significantly enhanced by a similar machine-learning approach.
Extending the understanding of Bose-Einstein condensate (BEC) physics to new geometries and topologies has a long and varied history in ultracold atomic physics. One such new geometry is that of a bubble, where a condensate would be confined to the surface of an ellipsoidal shell. Study of this geometry would give insight into new collective modes, self-interference effects, topology-dependent vortex behavior, dimensionality crossovers from thick to thin shells, and the properties of condensates pushed into the ultradilute limit. Here we propose to implement a realistic experimental framework for generating shell-geometry BEC using radiofrequency dressing of magnetically trapped samples. Such a tantalizing state of matter is inaccessible terrestrially due to the distorting effect of gravity on experimentally feasible shell potentials. The debut of an orbital BEC machine (NASA Cold Atom Laboratory, aboard the International Space Station) has enabled the operation of quantum-gas experiments in a regime of perpetual freefall, and thus has permitted the planning of microgravity shell-geometry BEC experiments. We discuss specific experimental configurations, applicable inhomogeneities and other experimental challenges, and outline potential experiments.
Atom chips are an excellent tool for studying ultracold degenerate quantum gases, due to the high degree of controllability afforded by the precise potentials generated from the current-carrying microfabricated wires on the chip surface. The geometries of the trapping potentials are inherently capable of realising extreme aspect ratios, and therefore creating model systems with effectively reduced dimensionality, particularly the theoretically-tractable one-dimensional Bose gas. In addition, the temporal tunability makes it possible to impart non-adiabatic changes on the trapping potentials, allowing experimental investigation of samples which have been brought out of equilibrium - a situation which is not fully theoretically understood. This thesis describes the implementation, development and characterisation of an experimental system for producing the first Bose-Einstein condensates of atomic rubidium 87 gas trapped on the surface of an atom chip in Nottingham. Such an apparatus is very complex and requires careful characterisation in order to run in a stable and reliable way. Details of the experimental setup are thoroughly outlined, including the vacuum system, lasers, electronics, computer control and timing, and the optical imaging system. A newly installed compact two-dimensional magneto-optical trap provides an loading rate of 5e7 atoms per second for loading a three-dimensional mirror-magneto-optical trap with 1.5e8 atoms, at a temperature of 300uK within 10s. The cloud is then sub-Doppler cooled to 50uK, and spin-polarised with 96% purity into the |F=2,mF=+2> ground state within 5ms, in preparation for loading a purely magnetic trap. A millimeter sized copper Z-shaped conductor located beneath the atom chip surface creates a Ioffe-Pritchard magnetic trap, into which the laser cooled cloud is loaded with 70% efficiency, and can be held with a vacuum-limited lifetime of 40s. Evaporative cooling then pre-cools the sample to below 20uK within 10s, to allow the subsequent loading into potentials created by the atom chip with 100% efficiency. A final evaporation stage then cools the cloud below the phase transition temperature of 800nK, resulting finally in pure BECs with 105 atoms confined using the atom chip. Key measurements of various properties of the trapped condensates are presented, which are important in order to characterise the system fully, and to compare with theoretical expectations. In particular, included are the variation of condensate fraction with temperature, the BEC expansion dynamics, and the condensate lifetime in the trap, for example. Finally, it is demonstrated how BECs can be produced on the atom chip without the use of external macroscopic coils, achieved by using novel, integrated sheet structures located beneath the chip surface - unique to this experimental system - to create the necessary bias fields.