SEARCH · Search NASA
Results for “evolutionary algorithm”
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.
Crystalline phase transitions and vibrational spectra of silicon up to multiterapascal pressures
A composite high-pressure phase diagram for silicon has been predicted for up to 4 TPa. Here, this diagram has been built using a combination of evolutionary algorithm-based structure searches, electronic density functional theory, lattice dynamics of perfect crystals, and anharmonic corrections for the solid state in conjunction with molecular dynamics for evaluating the melt curve. The anharmonic corrections to free energy, arising from both finite-temperature three-phonon interactions and temperature dependence of the lattice axial ratios, play a critical role in properly identifying the solid–solid transition boundaries of the orthorhombic structures. A double hexagonal close-packed structure has been found to be thermodynamically and dynamically stable, sandwiched between the experimentally observed base-centered orthorhombic and hexagonal close-packed structures. In addition, beyond 2.8 TPa, there exists a sequence of face-centered cubic-to-body-centered cubic-to-simple cubic transitions that are accompanied by the localization of electrons in the interstitial spaces between the ions. Supplementing the structural calculations, second- and third-order interatomic force constants were evaluated to compute the phonon vibration modes and linewidths, respectively. This allowed an elaborate analysis of the Raman and infrared spectra for all of the structures of silicon identified so far.
Fourier Analysis-Based Evolutionary Multi-Objective Multiphysics Optimization of Liquid-Cooled Heat Sinks
Optimal heat dissipation in power modules can significantly increase their power density. Removing the generated heat is critical for capturing the benefits of advanced semiconductor materials and improving the reliability of the device operation. This article proposes a design optimization method for liquid-cooled heat sinks that use a Fourier analysis-based tool and an evolutionary optimization algorithm to optimize the heat sink geometry for specified objectives. The optimized heat sink geometry is then compared with state-of-the-art solutions in literature based on finite element analysis of different designs. The proposed methodology can develop complex geometries that outperform the conventional heat sink geometries.
Plant Reload Optimization (prlo)
The PRLO framework is built on a modular and extensible architecture that tightly couples advanced evolutionary optimization algorithms with nuclear fuel depletion solvers (i.e., nuclear physics neutronics code). It supports exploring complex, high-dimensional design spaces constrained by user-specified operational, safety, and economic constraints. Objectives such as minimizing fresh fuel enrichment, flattening radial and axial power distributions, and maximizing discharge burnup are evaluated. PRLO’s equilibrium cycle optimization capability enables the identification of core configurations that maintain fuel cycle sustainability over extended planning horizons. Its integration with the RAVEN platform facilitates optimization of loading patterns or fuel shuffling schemes across multiple cycles. The interface with SIMULATE, a licensed industry-standard nodal code developed by Studsvik, ensures accurate neutronic and thermal-hydraulic feedback for reactor core design. PRLO’s automated workflow engine supports iterative design refinement, enabling utilities to streamline core design processes and meet evolving performance and regulatory targets.
Optics Studies for Multipass Energy Recovery at CEBAF: ER@CEBAF
Energy recovery linacs (ERLs), focus on recycling the kinetic energy of electron beam for the purpose of accelerating a newly injected beam within the same accelerating structure. The rising developments in the super conducting radio frequency technology, ERL technology has achieved several noteworthy milestones over the past few decades. In year 2003, Jefferson Lab has successfully demonstrated a single pass energy recovery at the CEBAF accelerator. Furthermore, they conducted successful experiments with IR-FEL demo and upgrades, as well as the UV FEL driver. This multi-pass, multi-GeV range energy recovery demonstration proposed to be carried out at CEBAF accelerator at Jefferson Lab focuses on demonstrating highest energy recovery in super conducting linac in the low-current range. Continuous electron beam accelerate up to 7.5 GeV within 5-passes and decelerate in the next 5-passes recovering RF energy and dumps at a low energy dump. The beamline optics design for recirculating linacs require special attention to avoid beam instabilities due to RF wakefields. Usually, multi-pass linac beam lines require stronger focusing at lower energies as that is necessary to avoid beam breakup (BBU) instabilities, even with this small beam current. The CEBAF linac optics optimization is focused on balancing over-focusing at higher energies and beta excursions at lower energies. The race-track-shaped geometry of CEBAF accelerator allows its linacs to accommodate multiple energy beams simultaneously, while individual recirculating arcs transporting one beam energy, are shared between accelerating/decelerating beams. For the linac optics optimization process, an extended strategy is used that is originally used in 6-pass Recirculating Linac design of the LHeC, to represent the ten passes through a single linac. Using proper mathematical expressions, linac optics optimization can be achieved with evolutionary genetic algorithms, with Multi-Objective optimization. This thesis introduces a CEBAF optics redesign tailored to accommodates the ER@CEBAF multi-pass ER scheme. The isochronous arcs were retuned to match into optics solutions for optimized 10-pass linacs. Within this work, a single bunch particle tracking analysis presented here focuses on the further improvements of the beamline and beam transportation.
Genetic algorithm-based optimisation of the few-group structure for lead fast reactors analysis
The optimal choice of the few-group structure for full-core transient analyses is still an open issue in reactor physics, especially for fast system like the lead fast reactor. One possible approach to select the group boundaries is represented by heuristic search algorithms, such as evolutionary ones. In this paper, a genetic algorithm coupled with the SIMMER code is employed to determine optimized six-group boundaries for the analysis of the ALFRED reactor. The Serpent Monte Carlo code is adopted to produce both the fine-group cross section library and the fine-group flux, used as a figure of merit to drive the genetic optimisation. The results show that the algorithm is indeed able to find satisfactory solutions that comply with the set objectives and can be reasonably interpreted in light of the underlying physics of the considered core. (authors)
Ethane and methane at high pressures: Structure and stability
We have performed a combined experimental and theoretical study of ethane and methane at high pressures of up to 120 GPa at 300 K using x-ray diffraction and Raman spectroscopies and the USPEX ab initio evolutionary structural search algorithm, respectively. For ethane, we have determined the crystallization point, for room temperature, at 2.7 GPa and also the low pressure crystal structure (phase A). This crystal structure is orientationally disordered (plastic phase) and deviates from the known crystal structures for ethane at low temperatures. Moreover, a pressure induced phase transition has been identified, for the first time, at 13.6 GPa to a monoclinic phase B, the structure of which is solved based on good agreement with the experimental results and theoretical predictions. For methane, our x-ray diffraction measurements are in agreement with the previously reported high-pressure structures and equation of state (EOS). Finally, we have determined the EOSs of ethane and methane, which provides a solid basis for the discussion of their relative stability at high pressures.
Theoretical Investigation of Delafossite-Cu 2 ZnSnO 4 as a Promising Photovoltaic Absorber
In the quest for efficient and cost-effective photovoltaic absorber materials beyond silicon, considerable attention has been directed toward exploring alternatives. One such material, zincblende-derived Cu2ZnSnS4 (CZTS), has shown promise due to its ideal band gap size and high absorption coefficient. However, challenges such as structural defects and secondary phase formation have hindered its development. In this study, we examine the potential of another compound, Cu2ZnSnO4 (CZTO), with a similar composition to CZTS as a promising alternative. Employing ab initio density function theory (DFT) calculations in combination with an evolutionary structure prediction algorithm, we identify that the crystalline phase of delafossite structure is the most stable among the 900 (meta)stable CZTO. Its thermodynamic stability at room temperature is also confirmed by the molecular dynamics study. Excitingly, this new phase of CZTO displays a direct band gap where the dipole-allowed transition occurs, making it a strong candidate for efficient light absorptions. Furthermore, the estimation of spectroscopic limited maximum efficiency (SLME) directly demonstrates the high potential of delafossite-CZTO as a photovoltaic absorber. Our numerical results suggest that delafossite-CZTO holds promise for future photovoltaic applications.
ECNet is an evolutionary context-integrated deep learning framework for protein engineering
Abstract Machine learning has been increasingly used for protein engineering. However, because the general sequence contexts they capture are not specific to the protein being engineered, the accuracy of existing machine learning algorithms is rather limited. Here, we report ECNet (evolutionary context-integrated neural network), a deep-learning algorithm that exploits evolutionary contexts to predict functional fitness for protein engineering. This algorithm integrates local evolutionary context from homologous sequences that explicitly model residue-residue epistasis for the protein of interest with the global evolutionary context that encodes rich semantic and structural features from the enormous protein sequence universe. As such, it enables accurate mapping from sequence to function and provides generalization from low-order mutants to higher-order mutants. We show that ECNet predicts the sequence-function relationship more accurately as compared to existing machine learning algorithms by using ~50 deep mutational scanning and random mutagenesis datasets. Moreover, we used ECNet to guide the engineering of TEM-1 β-lactamase and identified variants with improved ampicillin resistance with high success rates.
An adaptive granular representative volume element model with an evolutionary periodic boundary for hierarchical multiscale analysis
The hierarchical multiscale analysis normally utilizes a microscopic representative volume element (RVE) model to capture path/history-dependent macroscopic responses instead of using phenomenological constitutive models. However, for problems involving large deformation, the current RVE model used in geomechanics may lose representative properties due to the progressive distortion of the RVE box, unless a proper reinitialization is applied. This work develops an adaptive RVE model in conjunction with an evolutionary periodic boundary (EPB) algorithm for hierarchical multiscale analysis of granular materials undergoing large deformation based on a recent RVE model proposed for coupling molecular dynamics and the material point method. The proposed adaptive RVE model avoids the reinitialization of the RVE box that even undergoes extremely large shear deformation; meanwhile, it accounts for the deformation history of the RVE model and treats the interaction between boundary particles and other image particles in a more efficient way. Numerical examples with extremely large deformation are used to illustrate the adaptive granular RVE model enhanced by the proposed EPB algorithm. Furthermore, some key features of this new methodology are further discussed for clarification.
Deep Reinforcement Learning Based Control of Wind Turbines for Fast Frequency Response
In order to fulfill vital auxiliary grid services, such as load regulation, spin and non-spin reserve provision, and frequency support during emergencies, there is often a requirement for certain wind farms to operate in de-loaded modes. Leveraging the swift response capabilities of wind farms, this study demonstrates that reserving power in de-loaded modes can significantly enhance power grid stability and reliability during system contingencies. Controlling wind farms optimally for frequency support is intricate due to the nonlinearity of models and controllers and the complexity of wind farm interactions with power systems. Here, to address this challenge, this paper introduces a novel approach that integrates wind turbines into reinforcement learning-based solutions for frequency response. This innovative methodology utilizes the state-of-the-art reinforcement learning algorithm known as the surrogate-gradient-based evolutionary strategy. The proposed learning-based algorithm provides continuous control of wind farm output to rapidly stabilize system frequency and prevent unnecessary trips of under-frequency load shedding relays. To facilitate efficient training, parallel computing techniques are employed. The proposed methodology is evaluated on a modified IEEE-39 bus system, and simulation results reveal its efficacy in reliably supporting power system frequency and preventing the need for unnecessary load shedding.
Reinforcement learning in discrete action space applied to inverse defect design
Abstract Reinforcement learning (RL) algorithms that include Monte Carlo Tree Search (MCTS) have found tremendous success in computer games such as Go, Shiga and Chess. Such learning algorithms have demonstrated super-human capabilities in navigating through an exhaustive discrete action search space. Motivated by their success in computer games, we demonstrate that RL can be applied to inverse materials design problems. We deploy RL for a representative case of the optimal atomic scale inverse design of extended defects via rearrangement of chalcogen (e.g. S) vacancies in 2D transition metal dichalcogenides (e.g. MoS 2 ). These defect rearrangements and their dynamics are important from the perspective of tunable phase transition in 2D materials i.e. 2H (semi-conducting) to 1T (metallic) in MoS 2 . We demonstrate the ability of MCTS interfaced with a reactive molecular dynamics simulator to efficiently sample the defect phase space and perform inverse design—starting from randomly distributed S vacancies, the optimal defect rearrangement of defects corresponds a line defect of S vacancies. We compare MCTS performance with evolutionary optimization i.e. genetic algorithms and show that MCTS converges to a better optimal solution (lower objective) and in fewer evaluations compared to GA. We also comprehensively evaluate and discuss the effect of MCTS hyperparameters on the convergence to solution. Overall, our study demonstrates the effectives of using RL approaches that operate in discrete action space for inverse defect design problems.
Density-driven ionization potential depression and its effects on material properties
Changes in ionization potential at high density drive changes in material transport properties and often in crystal structure and equation of state. Current models used to describe and predict these effects have not been tested in regimes much beyond ~200 GPa, because of limitations in experimental techniques. Now, however, high-energy laser facilities with flexible pulse-shaping capabilities have made it possible to reach pressures in the 1000 GPa range, while maintaining temperatures below 1 eV. We have used these new methods to ramp-compress cobalt and magnesium and have measured shifts in the k-shell emission lines to infer changes in ionization stage, and x-ray diffraction to measure changes in crystal structure. In the case of cobalt, commonly-used detailed models for effective ionization based on isolated-atom electronic structure are shown to give the wrong answer, while self-consistent density-functional-theory-based models reproduce the data well. In the case of magnesium, zero-kelvin density functional theory predictions of density-driven electron localization in lattice interstitials (so-called electride phases) at high pressure are found to be too simple to describe material evolution along a laser ramp-compression pathway. Instead, by a combination of experiment, first-principles theory and evolutionary structure-searching algorithms, we have discovered several unexpected lower-symmetry metastable structures. First principles results find that the density of localized interstitial charge in the high-pressure magnesium phases is too low to be measurable with x-ray diffraction, and the body of theory work which has predicted high pressure electride phases for many systems, based on analysis of electron localization function, needs to be re-examined.
DEVELOPMENT AND APPLICATION OF RISK ANALYSIS TOOLKIT FOR PLANT RESOURCE OPTIMIZATION
This paper presents the development of methods and tools that are being designed to optimize plant operations (e.g., maintenance/replacement schedules and optimal maintenance postures for plant components) in a manner that is more cost effective than current approaches and makes better use of available component health and cost data. These methods include both data- and model-based optimization methods. Model-based optimization methods directly include reliability and cost models to determine an optimal plant operational strategy. We consider gradient-based and evolutionary (based on genetic algorithms) optimization methods. The second class of methods target more specific use cases (e.g., project schedule optimization) and are not based on reliability models directly, but they require specific component reliability and cost data. This class of methods is based on variants of the knapsack problem with an aim to determine an optimal project schedule that maximizes the overall NPV. This paper also presents multi-objective methods designed to identify an optimal maintenance posture based on a Pareto frontier analysis. Rather than dictating the “right” tradeoff (i.e., identify the absolute best posture), we show how it is possible to perform a trade space exploration approach (i.e., identify value and costs of several postures and let the analysis account for desired value and cost metrics). This is performed by identifying maintenance postures that maximize value (e.g., system availability) and minimize operational costs, i.e., the Pareto frontier in a value-cost trade space. For all these methods we present detailed applicative examples that show their validity from a decision-making perspective.
Determining best practices for using genetic algorithms in molecular discovery
Genetic algorithms (GAs) are a powerful tool to search large chemical spaces for inverse molecular design. However, GAs have multiple hyperparameters that have not been thoroughly investigated for chemical space searches. In this tutorial, we examine the general effects of a number of hyperparameters, such as population size, elitism rate, selection method, mutation rate, and convergence criteria, on key GA performance metrics. Here, we show that using a self-termination method with a minimum Spearman’s rank correlation coefficient of 0.8 between generations maintained for 50 consecutive generations along with a population size of 32, a 50% elitism rate, three-way tournament selection, and a 40% mutation rate provides the best balance of finding the overall champion, maintaining good coverage of elite targets, and improving relative speedup for general use in molecular design GAs.
Neural architecture search via similarity adaptive guidance
Evolutionary neural network architecture search (ENAS) has attracted the attention of many experts due to its global optimization capabilities to automatically search for convolutional neural network architectures based on the target task. The current search space for ENAS is not to design a fully structured network, but to search for smaller cell architectures to reduce search costs. However, blind search strategies do not effectively utilize the potential experience of the population. In order to utilize the potential experience learned by the current population to guide the evolutionary search of the population, we propose a similarity guided neural network architecture search algorithm based on cell architecture, which utilizes the similarity between pairwise architectures in the population as empirical knowledge learned by the population. Our proposed algorithm provides a novel method for calculating architecture similarity, which calculates architecture similarity separately from the cell and macro-structure. Then we decouple the connections and operations in the cell and calculate connection and operation similarity separately. In addition, we propose adaptive similarity selection and binary tournament selection strategies to enhance the algorithm’s global and local search capabilities and effectively explore the search space. Finally, we design an improved single-point crossover operator to enhance the local search ability of the evolutionary operator. The experimental results show that SAGNAS is a competitive algorithm that achieves 97.44% and 81.60% in CIFAR10 and CIFAR100 with only 1.9 GPU-days spent.
Visualization System for Evolutionary Neural Networks for Deep Learning
Deep learning is actively used in a wide range of fields for scientific discovery. To effectively apply deep learning to a particular problem, it is important to select an appropriate network architecture and other hyper-parameters (at each layer). Evolving architectures and hyper-parameters using a genetic algorithm is one current approach to search the huge space of all possible configurations to find those more optimal for the problem. However, examining an evolutionary process and tuning the genetic algorithm are challenging, pushing most users to treat the process as a black box. To address this challenge, we propose a visualization system for evolutionary neural networks for deep learning. The key feature of our visualization system is to provide a visual analytics environment for evaluating a genetic algorithm in order to improve the underlying operations to reduce time to find good solutions. Our system is able to not only visualize how a genetic algorithm traverses its search space but also allows users to examine evolving networks in-depth to get insights to improve performance through interactive visualization components.