Search NASA⌕ Search

SEARCH · Search NASA

Results for “evolutionary algorithms”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 127 records · Page 7

High pressure-induced distortion in face-centered cubic phase of thallium

The complex and unusual high-pressure phase transition of III-A (i.e. Al, Ga, and In) metals have been investigated in the last several decades because of their interesting periodic table position between the elements having metallic and covalent bonding. Here, our present first principles-based electronic structure calculations and experimental investigation have revealed the unusual distortion in face-centered cubic (f.c.c.) phase of the heavy element thallium (Tl) induced by the high pressure. We have predicted body-centered tetragonal (b.c.t) phase at 83 GPa using an evolutionary algorithm coupled with ab initio calculations, and this prediction has been confirmed with a slightly distorted parameter (2–√ × a-c)/c lowered by 1% using an angle-dispersive X-ray diffraction technique. The density functional theory (DFT)-based calculations suggest that s–p mixing states and the valence-core overlapping of 6s and 5d states play the most important roles for the phase transitions along the pathway h.c.p→f.c.c.→b.c.t.

36 MATERIALS SCIENCE↗

Genetic programming for the nuclear many-body problem: a guide

Genetic Programming (GP) is an evolutionary algorithm that generates computer programs, or mathematical expressions, to solve complex problems. In this Guide, we demonstrate how to use GP to develop surrogate models to mitigate the computational costs of modeling atomic nuclei with ever increasing complexity. The computational burden escalates when uncertainty quantification is pursued, or when observables must be globally computed for thousands of nuclei. By studying three models in which the mean field depends on the total particle density self-consistently, we show that by constructing reduced order models supported by GP one can speed up many-body computations by several orders of magnitude with a negligible loss in accuracy.

dimensionality reduction↗

Harnessing the power of gradient-based simulations for multi-objective optimization in particle accelerators

Abstract Particle accelerator operation requires simultaneous optimization of multiple objectives. Multi-objective optimization (MOO) is particularly challenging due to trade-offs between the objectives. Evolutionary algorithms, such as genetic algorithms (GAs), have been leveraged for many optimization problems, however, they do not apply to complex control problems by design. This paper demonstrates the power of differentiability for solving MOO problems in particle accelerators using a deep differentiable reinforcement learning (DDRL) algorithm. We compare the DDRL algorithm with model-free reinforcement learning (MFRL), GA, and Bayesian optimization (BO) for simultaneous optimization of heat load and trip rates in the continuous electron beam accelerator facility. The underlying problem enforces strict constraints on both individual states and actions as well as cumulative (global) constraints on energy requirements of the beam. Using historical accelerator data, we develop a physics-based surrogate model which is differentiable and allows for back-propagation of gradients. The results are evaluated in the form of a Pareto-front with two objectives. We show that the DDRL outperforms MFRL, BO, and GA on high dimensional problems.

43 PARTICLE ACCELERATORS↗

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

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 {approx}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.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Light storage in a doped solid enhanced by feedback-controlled pulse shaping

We report on experiments dealing with feedback-controlled pulse shaping to optimize the efficiency of light storage by electromagnetically induced transparency (EIT) in a Pr{sup 3+}:Y{sub 2}SiO{sub 5} crystal. A learning loop in combination with an evolutionary algorithm permits the automatic determination of optimal temporal profiles of intensities and frequencies in the driving laser pulses (i.e., the probe and coupling pulses). As a main advantage, the technique finds optimal solutions even in the complicated multilevel excitation scheme of a doped solid, involving large inhomogeneous broadening. The learning loop experimentally determines optimal temporal intensity profiles of the coupling pulses for a given probe pulse. The optimized intensity pulse shapes enhance the light-storage efficiency in the doped solid by a factor of 2. The learning loop also determines a fast and efficient preparation pulse sequence, which serves to optically prepare the crystal prior to light-storage experiments. The optimized preparation sequence is 5 times faster than standard preparation sequences. Moreover, the optimized preparation sequence enhances the optical depth in the medium by a factor of 5. As a consequence, the efficiency of light storage also increases by another factor of 3. Our experimental data clearly demonstrate the considerable potential of feedback-controlled pulse shaping, applied to EIT-driven light storage in solid media.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Phase stability in a multistage Zeeman decelerator

The phase stability of a multistage Zeeman decelerator is analyzed by numerical particle-trajectory simulations and experimental measurements. A one-dimensional model of the phase stability in multistage Stark deceleration [Bethlem et al., Phys. Rev. Lett. 84, 5744 (2000)] has been adapted to multistage Zeeman deceleration and compared with one- and three-dimensional particle-trajectory simulations, including the analysis of the effect of finite switch-on and -off times of the deceleration pulses. The comparison reveals that transverse effects in the decelerator lead to a considerable reduction of the phase-space acceptance at low values of the phase angle and an enhancement at high values. The optimal combinations of phase angles and currents with which a preset amount of kinetic energy can be removed from atoms and molecules in a pulsed supersonic beam using a multistage decelerator are determined by simulation. Quantitative analysis of the phase-space acceptance within a given volume reveals that for our decelerator (8 {mu}s switch-off time) optimal conditions are achieved for values of the phase angle between 45 deg. and 55 deg. This conclusion is examined and confirmed by experimental measurements using deuterium atoms. Alternative approaches to generate optimal deceleration pulse sequences, such as the implementation of evolutionary algorithms or the use of higher-order modes of the decelerator, are discussed.

74 ATOMIC AND MOLECULAR PHYSICS↗

Evolutionary optimization of rotational population transfer

We present experimental and numerical studies on control of rotational population transfer of NO(J=1/2) molecules to higher rotational states. We are able to transfer 57% of the population to the J=5/2 state and 46% to J=9/2, in good agreement with quantum mechanical simulations. The optimal pulse shapes are composed of pulse sequences with delays corresponding to the beat frequencies of states on the rotational ladder. The evolutionary algorithm is limited by experimental constraints such as volume averaging and the finite laser intensity used, the latter to circumvent ionization. Without these constraints, near-perfect control (>98%) is possible. In addition, we show that downward control, moving molecules from high to low rotational states, is also possible.

74 ATOMIC AND MOLECULAR PHYSICS↗

Predicting the ground-state structure of sodium boride

Binary borides has been a subject of extensive research. However, the exact compositions and crystal structures of sodium borides remained controversial. Furthermore, using the ab initio variable-composition evolutionary algorithm, a new stable Na 2 B 30 with I2 1 2 1 2 1 symmetry (I2 1 2 1 2 1 –Na 2 B 30 ) is found, which is –7.38 meV / atom lower in energy than the Imma–Na 2 B 30 structure reported by experimentalists. Interestingly, the Imma–Na 2 B 30 structure is predicted to be a topological nodal line semimetal, which may result in superior electronic transport. In contrast, I2 1 2 1 2 1 –Na 2 B 30 is an ultrahard semiconductor with an unprecedented open-framework structure, whose interstitial helical boron sublattice enhances its hardness and energetic stability.

36 MATERIALS SCIENCE↗

Magnetic borophenes from an evolutionary search

A computational methodology based on ab initio evolutionary algorithms and spin-polarized density functional theory was developed to predict two-dimensional magnetic materials. Its application to a model system borophene reveals an unexpected rich magnetism and polymorphism. A metastable borophene with nonzero thickness is an antiferromagnetic semiconductor from first-principles calculations, and can be further tuned into a half-metal by finite electron doping. In this borophene, the buckling and coupling among three atomic layers are not only responsible for magnetism, but also result in an out-of-plane negative Poisson's ratio under uniaxial tension, making it the first elemental material possessing auxetic and magnetic properties simultaneously.

2-dimensional systems↗

Model-independent analysis of Airy structures in the {sup 16}O+{sup 12}C and {sup 16}O+{sup 16}O elastic scattering differential cross sections at 13-22 MeV/nucleon

We present the results of the model-independent analysis of Airy structures in the {sup 16}O+{sup 12}C and {sup 16}O+{sup 16}O elastic scattering differential cross sections at 13-22 MeV/nucleon. The analysis has been performed with help of a procedure based on the application of the evolutionary algorithm, which enables us to extract the nuclear part of the scattering matrix S{sub N}(l) as a complex function of angular momentum directly from the scattering data. Contrary to the commonly used model approaches, our procedure gives the better fits and leads to the S{sub N}(l) representations defined by the moduli and the nuclear phases exhibiting smooth monotonic dependencies on l.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Anharmonic and Anomalous Trends in the High-Pressure Phase Diagram of Silicon

A multifaceted first-principles approach utilizing density functional theory, evolutionary algorithms, and lattice dynamics was used to construct the phase diagram of silicon up to 4 TPa and 26,000 K. These calculations predicted that (1) an anomalous sequence of face-centered cubic (fcc) to body-centered cubic (bcc) to simple cubic (sc) crystalline phase transitions occur at pressures of 2.87 TPa and 3.89 TPa, respectively, along the cold curve; (2) the orthorhombic phases of Imma and Cmce-16 appear on the phase diagram only when the anharmonic contribution to the Gibbs free energy is taken into account; and (3) a drastic change in the slope of the principal Hugoniot is observed if the negative thermal expansion coefficient of the cubic diamond phase is taken into account through anharmonic effects.

36 MATERIALS SCIENCE↗

Study for material analogs of FeSb 2 : Material design for thermoelectric materials

Using the ab initio evolutionary algorithm (implemented in uspex) and electronic structure calculations we investigate the properties of a new thermoelectric material FeSbAs, which is a material analog of the enigmatic thermoelectric FeSb 2 . We utilize the density functional theory and the Gutzwiller method to check the energetics. We find that FeSbAs can be made thermodynamically stable above ~ 30 GPa. We investigate the electronic structure and thermoelectric properties of FeSbAs based on the density functional theory and compare with those of FeSb 2 . Above 50 K, FeSbAs has higher Seebeck coefficients than FeSb 2 . Upon doping, the figure of merit becomes larger for FeSbAs than for FeSb 2 . Another material analog FeSbP, was also investigated, and found thermodynamically unstable even at very high pressure. Regarding FeSb 2 as a member of a family of compounds (FeSb 2 , FeSbAs, and FeSbP) we elucidate what are the chemical handles that control the gaps in this series. Here, we also investigate solubility (As or P for Sb in FeSb 2 ) we found As to be more soluble. Finally, we study a two-band model for thermoelectric properties and find that the temperature dependent chemical potential and the presence of the ionized impurities are important to explain the extremum in the Seebeck coefficient exhibited in experiments for FeSb 2 .

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗

Two-dimensional Dirac semimetal based on the alkaline earth metal CaP 3

Using an evolutionary algorithm in combination with first-principles density-functional theory calculations, we identify a two-dimensional (2D) CaP 3 monolayer as a new Dirac semimetal due to inversion and nonsymmorphic spatial symmetries of the structure. This new topological material, composed of light elements, exhibits high structural stability (higher than the phase known in the literature), which is confirmed by thermodynamic and kinetic stability analysis. Moreover, it satisfies the electron filling criteria, so that its Dirac state is located near the Fermi level. The existence of the Dirac state predicted by the theoretical symmetry analysis is also confirmed by first-principles electronic band structure calculations. We find that the energy position of the Dirac state can be tuned by strain, while the Dirac state is unstable against an external electric field since it breaks the spatial inversion symmetry. In conclusion, our findings should be instrumental in the development of 2D Dirac fermions based on light elements for their application in nanoelectronic devices and topological electronics.

2-dimensional systems↗

A science-driven approach to optimize the design for a biological small-angle neutron scattering instrument

Biological small-angle neutron scattering (SANS) instruments facilitate critical analysis of the structure and dynamics of complex biological systems. However, with the growth of experimental demands and the advances in optical systems design, a new neutron optical concept is necessary to overcome the limitations of current instruments. This work presents an approach to include experimental objectives ( i.e. the science to be supported by a specific neutron scattering instrument) in the optimization of the neutron optical concept. The approach for a proposed SANS instrument at the Second Target Station of the Spallation Neutron Source at Oak Ridge National Laboratory, USA, is presented here. Further, the instrument is simulated with the McStas software package. The optimization process is driven by an evolutionary algorithm using McStas output data, which are processed to calculate an objective function designed to quantify the expected performance of the simulated neutron optical configuration for the intended purpose. Each McStas simulation covers the complete instrument, from source to detector, including realistic sample scattering functions. This approach effectively navigates a high-dimensional parameter space that is otherwise intractable; it allows the design of next-generation SANS instruments to address specific scientific cases and has the potential to increase instrument performance compared with traditional design approaches.

47 OTHER INSTRUMENTATION↗

Confocal depth-resolved fluorescence micro-X-ray absorption spectroscopy for the study of cultural heritage materials: a new mobile endstation at the Beijing Synchrotron Radiation Facility

A confocal fluorescence endstation for depth-resolved micro-X-ray absorption spectroscopy is described. A polycapillary half-lens defines the incident beam path and a second polycapillary half-lens at 90° defines the probe sample volume. An automatic alignment program based on an evolutionary algorithm is employed to make the alignment procedure efficient. This depth-resolved system was examined on a general X-ray absorption spectroscopy (XAS) beamline at the Beijing Synchrotron Radiation Facility. Sacrificial red glaze (AD 1368–1644) china was studied to show the capability of the instrument. As a mobile endstation to be applied on multiple beamlines, the confocal system can improve the function and flexibility of general XAS beamlines, and extend their capabilities to a wider user community.

59 BASIC BIOLOGICAL SCIENCES↗

Evolving Larger Convolutional Layer Kernel Sizes for a Settlement Detection Deep-Learner on Summit

Deep-learner hyper-parameters, such as kernel sizes, batch sizes, and learning rates, can significantly influence the quality of trained models. The state of the art for finding optimal hyper-parameters generally uses a brute force, grid search approach, random search, or Bayesian-based optimization among other techniques. We applied an evolutionary algorithm to optimize kernel sizes for a convolutional neural network used to detect settlements in satellite imagery. Usually convolutional layer kernel sizes are small - typically one, three, or five - but we found that the system converged at, or near, kernel sizes of nine for the last convolutional layer, and that this occurred for multiple runs using two different datasets. Moreover, the larger kernel sizes had fewer false positives than the 3x3 kernel sizes found as optimal via a brute force uniform grid search. This suggests that this large kernel size may be leveraging patterns found in larger areal features in the source imagery, and that this may be generalized as possible guidance for similar remote sensing deep-learning tasks.

Coletti, Mark↗

Predicting Band-Gap of Inorganic Materials Using Neuromorphic Graph Learning

Predicting properties of inorganic materials is a heavily researched topic, with several new prediction approaches emerging as competitors. One such competitor is graph neural networks, which leverage the structure of the graph to aid in the prediction process. In this work, we propose integration of neuromorphic computation into the graph neural network pipeline. We call this approach Neuromorphic Graph Learning (NGL). We utilize the NGL approach to leverage evolutionary algorithms and a novel Spike Pipeline for Raster Analysis (SPIRE) for the prediction of band gap in inorganic materials.

Mulet, Ian [University of Tennessee (UT)]↗

Evolving bipartite authentication graph partitions

As large scale enterprise computer networks become more ubiquitous, finding the appropriate balance between user convenience and user access control is an increasingly challenging proposition. Suboptimal partitioning of users’ access and available services contributes to the vulnerability of enterprise networks. Previous edge-cut partitioning methods unduly restrict users’ access to network resources. This paper introduces a novel method of network partitioning superior to the current state-of-the-art which minimizes user impact by providing alternate avenues for access that reduce vulnerability. Networks are modeled as bipartite authentication access graphs and a multi-objective evolutionary algorithm is used to simultaneously minimize the size of large connected components while minimizing overall restrictions on network users. Lastly, results are presented on a real world data set that demonstrate the effectiveness of the introduced method compared to previous naive methods.

97 MATHEMATICS AND COMPUTING↗