Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Genetic 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 469 records · Page 26

Design of High Power Density 100 kW Surface Permanent Magnet Machine with No Heavy Rare Earth Material Using Current Source Inverter for Traction Application

Surface permanent magnet (SPM) machines are appealing candidates for traction applications because of high power density and high efficiency. Rare-earth magnets deliver high performance but raise concerns about material supply dependability, particularly if they use heavy rare-earth materials. This paper presents the design of a high-performance SPM machine without any heavy rare-earth material that is optimized specifically for a current-source inverter (CSI) traction drive. This machine is designed for a constant-power speed ratio (CPSR) of 3. A genetic algorithm has been performed to achieve a target of 50 kW/L active power density at speeds up to 20,000 rpm. Operation at a peak line-to-line voltage of 800 V can be achieved using 1200 V power devices with a safety margin. The predicted electromagnetic and rotor structural performance characteristics are presented using both analytical and finite element analysis (FEA) results.

(CPSR)↗

An evolutionary algorithm for designing microbial communities via environmental modification

Despite a growing understanding of how environmental composition affects microbial communities, it remains difficult to apply this knowledge to the rational design of synthetic multispecies consortia. This is because natural microbial communities can harbour thousands of different organisms and environmental substrates, making up a vast combinatorial space that precludes exhaustive experimental testing and computational prediction. Here, we present a method based on the combination of machine learning and metabolic modelling that selects optimal environmental compositions to produce target community phenotypes. In this framework, dynamic flux balance analysis is used to model the growth of a community in candidate environments. A genetic algorithm is then used to evaluate the behaviour of the community relative to a target phenotype, and subsequently adjust the environment to allow the organisms to approach this target. We apply this iterative process to thousands of in silico communities of varying sizes, showing how it can rapidly identify environments that yield desired taxonomic compositions and patterns of metabolic exchange. Moreover, this combination of approaches produces testable predictions for the assembly of experimental microbial communities with specific properties and can facilitate rational environmental design processes for complex microbiomes.

59 BASIC BIOLOGICAL SCIENCES↗

Multi-agent voltage control in distribution systems using GAN-DRL-based approach

Active distribution grids can experience voltage fluctuations and violations due to the high penetration of variable distributed energy resources (DERs). These problems might occur because of the uncertain and variable generation natures of these resources, especially solar photovoltaic resources, during panel shadowing scenarios. Volt-VAR control (VVC) is an efficient method that controls the reactive power set-points of the inverters to regulate the voltage of distribution grids. Although several VVC approaches have been proposed recently, the performance of these approaches degrades significantly if behind-the-meter solar generation data are unobservable/missing. Therefore, it is necessary to impute missing/unobservable PV data accurately to be utilized in VVC approaches. Further, this paper proposes a model-free, data-driven, centrally trained, and decentrally executed multi-agent deep reinforcement learning-based VVC architecture to regulate the voltage of distribution networks. A generative adversarial network (GAN) is incorporated to impute the unobservable PV data accurately, which improves the performance of the proposed control architecture. The proposed multi-agent-soft-actor–critic algorithm (MASAC)-based VVC technique utilizes the actual PV dataset as well as the imputed dataset from the GAN framework to learn the optimal coordinated control policy for controlling the optimal reactive power set-points of PV inverters. The effectiveness of the proposed approach is analyzed on a modified IEEE 34-bus test case with added PV inverters. The results are compared and analyzed with a base case model with no VVC and VVC with a local droop control approach, genetic algorithm optimization, and a centralized soft actor–critic-based approach. Moreover, the performance of the proposed approach is compared with that of a multi-agent VVC framework without using the PV generation data and load information as the system state. The results illustrate that the proposed method with more state input improves the voltage profile and reduces the power loss of the network across various loading and PV generation scenarios.

14 SOLAR ENERGY↗

Application of Fireworks Algorithm in Gamma-Ray Spectrum Fitting for Radioisotope Identification

We present that identification of radioisotopic signature patterns in gamma-ray spectra is of paramount importance in various applications of gamma spectroscopy. Therefore, there are several active research efforts to develop accurate and precise methods to perform automated spectroscopic analysis and subsequently recognize gamma-ray signatures. In this work, the authors present a new method for radioisotope identification in gamma-ray spectra obtained with a low resolution radiation detector. The method fits the obtained spectrum with a linear combina-tion of known template signature patterns. Coefficients of the linear combination are evaluated by computing the solution of a single objective optimization problem, whose objective is the Theil-1 inequality coefficient. Optimization of the problem is performed by the Fireworks Algorithm, which identifies a set of coefficients that minimize the Theil-1 value. The computed coefficients are statistically tested for being significantly different than zero or not, and if at least one is found to be zero then the Fireworks Algorithm is used to reiterate fitting using the non-zero templates. Fitting iterations are continued up to the point that no linear coefficients are found to be zero. The output of the method is a list that contains the radioisotopes that have been identified in the measured spectrum. The method is tested on a set of both simulated and real experimental gamma-ray spectra comprised of a variety of isotopes, and compared to a multiple linear regression fitting, and genetic algorithm Theil-1 based fitting. Finally, results demonstrate the potentiality of the Fireworks Algorithm based method, expressed as higher accuracy and similar precision over the other two tested methodologies for radioisotope signature pattern identification in the framework of gamma-ray spectrum fitting.

Theil Coefficient↗

Optimizing deep learning hyper-parameters through an evolutionary algorithm. In: MLHPC '15 Proceedings of the Workshop on Machine Learning in High-Performance Computing Environments, Article No. 4

There has been a recent surge of success in utilizing Deep Learning (DL) in imaging and speech applications for its relatively automatic feature generation and, in particular for convolutional neural networks (CNNs), high accuracy classification abilities. While these models learn their parameters through data-driven methods, model selection (as architecture construction) through hyper-parameter choices remains a tedious and highly intuition driven task. To address this, Multi-node Evolutionary Neural Networks for Deep Learning (MENNDL) is proposed as a method for automating network selection on computational clusters through hyper-parameter optimization performed via genetic algorithms.

Young, Steven R.↗

Assessment of Gamma-Ray-Spectra Analysis Method Utilizing the Fireworks Algorithm for Various Error Measures

The analysis of measured data plays a significant role in enhancing nuclear nonproliferation mainly by inferring the presence of patterns associated with special nuclear materials. Among various types of measurements, gamma-ray spectra is the widest utilized type of data in nonproliferation applications. In this paper, a method that employs the fireworks algorithm (FWA) for analyzing gamma-ray spectra aiming at detecting gamma signatures is presented. In particular, FWA is utilized to fit a set of known signatures to a measured spectrum by optimizing an objective function, where non-zero coefficients express the detected signatures. FWA is tested on a set of experimentally obtained measurements optimizing various objective functions—MSE, RMSE, Theil-2, MAE, MAPE, MAP—with results exhibiting its potential in providing highly accurate and precise signature detection. Finally and furthermore, FWA is benchmarked against genetic algorithms and multiple linear regression, showing its superiority over those algorithms regarding precision with respect to MAE, MAPE, and MAP measures.

97 MATHEMATICS AND COMPUTING↗

An Evolutionary Algorithm Approach to Multi-Pass ERL Optics Design

An Energy Recovery Experiment at CEBAF (ER@CEBAF) is aimed at demonstrating high energy, low current, multi-pass energy recovery at the existing 12 GeV CEBAF accelerator. The beam break-up instability, limiting the maximum beam current, can be controlled through minimizing beta functions for the lowest energy pass, which gives a preference to strongly focusing optics, e.g. a semi-periodic FODO lattice. On the other hand, one needs to limit beta function excursions, caused by under focusing, at the higher energy passes, which in turn favors weakly focusing linac optics. Balancing both effects is the main objective of proposed multi-pass linac optics optimization. Here, we discuss an optics design process for ER@CEBAF transverse optics using a genetic algorithm.

Neththikumara, I.↗

Nonlinear Dynamic Model-Based Multiobjective Sensor Network Design Algorithm for a Plant with an Estimator-Based Control System

Here, a novel sensor network design (SND) algorithm is developed for maximizing process efficiency while minimizing sensor network cost for a nonlinear dynamic process with an estimator-based control system. The multiobjective optimization problem is solved following a lexicographic approach where the process efficiency is maximized first followed by minimization of the sensor network cost. The partial net present value, which combines the capital cost due to the sensor network and the operating cost due to deviation from the optimal efficiency, is proposed as an alternative objective. The unscented Kalman filter is considered as the nonlinear estimator. The large-scale combinatorial optimization problem is solved using a genetic algorithm. The developed SND algorithm is applied to an acid gas removal (AGR) unit as part of an integrated gasification combined cycle (IGCC) power plant with CO 2 capture. Due to the computational expense, a reduced order nonlinear model of the AGR process is identified and parallel computation is performed during implementation.

Lexicographic optimization↗

Bridging Hydrological Ensemble Simulation and Learning Using Deep Neural Operators

Ensemble-based simulation and learning (ESnL) has long been used in hydrology for parameter inference, but computational demands of process-based ESnL can be quite high. To address this issue, we propose a deep neural operator learning approach. Neural operators are generic machine learning algorithms that can learn functional mappings between infinite-dimensional spaces, providing a highly flexible tool for scientific machine learning. Our approach is built upon DeepONet, a specific deep neural operator, and is designed to address several common problems in hydrology, namely, model parameter estimation, prediction at ungaged locations, and uncertainty quantification. Here we demonstrate the effectiveness of our DeepONet-based workflow using an existing large model ensemble created for an eastern U.S. watershed that is instrumented with 10 streamflow gages. Results suggest DeepONet achieves high efficiency in learning an ML surrogate model from the model ensemble, with the modified Kling-Gupta Efficiency exceeding 0.9 on holdout test sets. Parameter inference, carried out using the trained DeepONet surrogate model and genetic algorithm, also yields robust results. Additionally, we formulate and train a separate DeepONet model for physics-informed, seq-to-seq streamflow forecasting, which further reduces biases in the pre-trained DeepONet surrogate model. While this study focuses primarily on a single watershed, our approach is general and may be extended to enable learning from model ensembles across multiple basins or models. Thus, this research represents a significant contribution to the application of hybrid machine learning in hydrology.

54 ENVIRONMENTAL SCIENCES↗

Event reconstruction for line source releases

The goal of source inversion, also called event reconstruction, is the calculation of source parameters from information obtained by network of concentration (or dosage) and meteorological sensors. Source parameters include source location and strength, but in certain cases there could be more than one source so the inversion procedure could deal with determination of number of sources, as well. In a case of limited time period pollutant emission events, as for example during accidents or intentional releases, it is of great use to estimate starting and ending times of the event. This kind of research is very useful for estimating the source parameters of industrial pollutants since it provides important information for regulation purposes. Also it provides information to fast responders in a case of accidental pollutant releases or for homeland security needs when chemical, biological or radiological agent is deliberately released. Development of faster and more accurate algorithms is very important since it could help reduce the populace's exposure to dangerous airborne contaminants, plan evacuation routes, and help assess the magnitude of clean up. During the last decade, the large number of research papers in area of source inversion was published where many different approaches were used. Most of the source inversion work publish to date apply to point source releases. The forward dispersion models used range from fast Gaussian plume and puff codes that enable almost instantaneous calculations of concentrations and dosages to Computational Fluid Dynamics (CFD) codes that provide more detailed and precise calculation but at the same time are expensive with respect to time and computer resources. The optimization methods were often used and examples are simulated annealing and genetic algorithms.

45 MILITARY TECHNOLOGY, WEAPONRY, AND NATIONAL DEF↗

Quantifying uncertainty for deep learning based forecasting and flow-reconstruction using neural architecture search ensembles

Classical problems in computational physics such as data-driven forecasting and signal reconstruction from sparse sensors have recently seen an explosion in deep neural network (DNN) based algorithmic approaches. However, most DNN models do not provide uncertainty estimates, which are crucial for establishing the trustworthiness of these techniques in downstream decision making tasks and scenarios. In recent years, ensemble-based methods have achieved significant success for the uncertainty quantification in DNNs on a number of benchmark problems. However, their performance on real-world applications remains under-explored. In this work, we present an automated approach to DNN discovery and demonstrate how this may also be utilized for ensemble-based uncertainty quantification. Specifically, we propose the use of a scalable neural and hyperparameter architecture search for discovering an ensemble of DNN models for complex dynamical systems. We highlight how the proposed method not only discovers high-performing neural network ensembles for our tasks, but also quantifies uncertainty seamlessly. This is achieved by using genetic algorithms and Bayesian optimization for sampling the search space of neural network architectures and hyperparameters. Subsequently, a model selection approach is used to identify candidate models for an ensemble set construction. Afterwards, a variance decomposition approach is used to estimate the uncertainty of the predictions from the ensemble. We demonstrate the feasibility of this framework for two tasks — forecasting from historical data and flow reconstruction from sparse sensors for the sea-surface temperature. In conclusion, we demonstrate superior performance from the ensemble in contrast with individual high-performing models and other benchmarks.

Deep ensembles↗

Towards an Agent-Based Blackboard System for Reactor Design Optimization

The field of reactor design is rich with opportunities for applications of computational optimization algorithms; these applications can range from preliminary core design to reactor shuffling patterns. Many of these schemes rely on sets of previously generated solutions (sometimes referred to as “generations”) to inform future decisions. While it is important to build upon prior knowledge, this process requires a full generation of solutions to be formed before future solutions can be examined. Rather than relying on a generational scheme to perform an optimization, we propose using an agent-based approach in conjunction with a blackboard framework for performing reactor design optimizations. Utilizing an agent-based approach allows agents to perform tasks independently, while retaining the ability to build off of previous solutions. We develop an agent-based blackboard system (ABBS) for determining the Pareto front (PF) in sodium fast reactor design optimization problems and compared this with the Non-Dominated Sorting Genetic Algorithm II (NSGA-II). Our goal is to evaluate the viability of the ABBS in producing a PF that is comparable with the NSGA-II algorithm. The design space consists of the fuel height, fuel smear, and plutonium fraction in the core, and we seek to minimize the reactivity swing and plutonium mass, while maximizing the burnup. The diversity, coverage, and spread of the PFs generated by the two methods are examined, and the ABBS is able to converge to the same PF as the NSGA-II algorithm. These results show that the ABBS is able to find optimal designs that are similar to those found by the NSGA-II algorithm. We conclude our study by applying the ABBS to the design of a sodium-cooled fast reactor to dispose of weapons-grade plutonium. The ABBS finds a core design that can burn upwards of 17.5 kg of weapons-grade plutonium per year and degrade an additional 195 kg of weapons-grade plutonium per year into non-weapons-grade material.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

Performance Portability of Molecular Docking Miniapp On Leadership Computing Platforms

Rapidly changing computer architectures, such as those found at high-performance computing (HPC) facilities, present the need for mini-applications (miniapps) that capture essential algorithms used in large applications to test program performance and portability, aiding transitions to new systems. The COVID-19 pandemic has fueled a flurry of activity in computational drug discovery, including the use of supercomputers and GPU acceleration for massive virtual screens for therapeutics. Recent work targeting COVID-19 at the Oak Ridge Leadership Computing Facility (OLCF) used the GPU-accelerated program AutoDock-GPU to screen billions of compounds on the Summit supercomputer. In this paper we present the development of a new miniapp, miniAutoDock-GPU, that can be used to evaluate the performance and portability of GPU-accelerated protein-ligand docking programs on different computer architectures. These tests are especially relevant as facilities transition from petascale systems and prepare for upcoming exascale systems that will use a variety of GPU vendors. The key calculations, namely, the Lamarckian genetic algorithm combined with a local search using a Solis-Wets based random optimization algorithm, are implemented. We developed versions of the miniapp using several different programming models for GPU acceleration, including a version using the CUDA runtime API for NVIDIA GPUs, and the Kokkos middle-ware API which is facilitated by C++ template libraries. A third version, currently in progress, uses the HIP programming model. These efforts will help facilitate the transition to exascale systems for this important emerging HPC application, as well as its use on a wide range of heterogeneous platforms.

Thavappiragasam, Mathialakan↗

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.

42 ENGINEERING↗

Quantum control for initiation and detection of explosives

We employ quantum control methods towards detection and quantum controlled initiation (QCI) of energetic materials. Ultrafast pulse shaping of broadband Infrared ({approx}750 nm to 850 run) and ultraviolet (266 nm, 400 nm) light is utilized for control. The underlying principals behind optimal control can be utilized to both detect and initiate explosives. In each case, time dependent phase shaped electric fields drive the chemical systems towards a desired state. For optimal dynamic detection of explosives (ODD-Ex) a phase specific broadband infrared pulse is created which increases not only the sensitivity of detection but also the selectivity of an explosive's spectral signatures in a background of interferents. QCI on the other hand, seeks to initiate explosives by employing shaped ultraviolet light. QCI is ideal for use with explosive detonators as it removes the possibility of unintentional initiation from an electrical source while adding an additional safety feature, initiation only with the proper pulse shape. Quantum control experiments require: (1) the ability to phase and amplitude shape the laser pulse and (2) the ability to effectively search for the pulse shape which controls the reaction. In these adaptive experiments we utilize both global and local optimization search routines such as genetic algorithm, differential evolution, and downhill simplex. Pulse shaping the broadband IR light, produced by focusing 800 nm light through a pressurized tube of Argon, is straightforward as commercial pulse shapers are available at and around 800 nm. Pulse shaping in the UV requires a home built shaper. Our system is an acoustic optical modulator (AOM) pulse shaper in which consists of a fused silica AOM crystal placed in the Fourier plane of a 4-f zero dispersion compressor.

66↗

Application of a convolutional neural network in permeability prediction: A case study in the Jacksonburg-Stringtown oil field, West Virginia, USA

Permeability is a critical parameter for understanding subsurface fluid flow behavior, managing reservoirs, enhancing hydrocarbon recovery, and sequestering carbon dioxide. In general, permeability is measured in the laboratory based on subsurface core samples, calculated from well logs or estimated from well tests. However, laboratory measurements and well tests are expensive, time-consuming, and usually limited to a few core samples or wells in a hydrocarbon field or carbon storage site. Machine-learning techniques are good options for generating a rapid, robust, and cost-effective permeability prediction model because of their strengths to recognize the potential interrelationships between input and output variables. Convolutional neural networks (CNN), as a good pattern recognition algorithm, are widely used in image processing, natural language processing, and speech recognition, but are rarely used with regression problems and even less often in reservoir characterization. We have developed a CNN regression model to estimate the permeability in the Jacksonburg-Stringtown oil field, West Virginia, which is a potential carbon storage site and enhanced oil recovery operations field. We also evaluate the concept of the geologic feature image, which is converted from geophysical well logs. Five variables, including two commonly available conventional well logs (the gamma rays [GRs] and bulk density) and three well-log-derived variables (the slopes of the GR and bulk density curves, and shale content), are used to generate a geologic feature image. The CNN treats the geologic feature image as the input and the permeability as the desired output. In addition, the permeability predicted using traditional backpropagation artificial neural networks, which are optimized by genetic algorithms and particle swarm optimization, is compared with the permeability estimated using our CNN. Our results indicate that the CNN regression model provides more accurate permeability predictions than the traditional neural network.

Geochemistry & Geophysics↗

A cloud computing approach to massively-parallel distributed stochastic optimization with application to geomechanical reservoir characterization

Many search strategies used to optimize parameters during hydrologic model calibration are inherently limited in that simulations must be performed sequentially in order for the algorithm to converge. As a result, the overall time required to reach a solution increases linearly with the number of steps necessary. In contrast, sampling strategies that attempt to approximate the posterior distribution of the parameters do not require the model simulations to be run in sequence and therefore invite the use of a massively parallel approach across heterogeneous computing environments. We apply a cloud computing methodology to this problem where a head node running on Amazon Web Services continuously produces candidate models that are distributed to arbitrary compute environments to run the simulation, in our case Clemson’s Palmetto cluster computer. Thus the time required to obtain a sufficient number of samples to produce an adequate approximation of the posterior distribution of the parameters is limited by the number of available compute sites. Monte Carlo sampling of the posterior space would be duplicative and inefficient, thus we combine Latin hypercube methods with genetic algorithms to ensure that no large regions of the parameter space are left unexplored, and that high probability regions are investigated thoroughly. We independently assess the fit to different types of data sets without the need to construct a single objective function. Sets of Pareto optimal models are particularly valuable when the quality and information content of data is poorly known or conceptual model errors exist in our simulation. We illustrate our approach for a problem where in situ strain data are used to calibrate the parameters of an oil reservoir in the Avant oil field in Oklahoma. We use well injection and shut-in to stress a deep (500m) confined formation, and observe the geomechanical response using borehole strainmeters installed in the shallow subsurface (30m). These data were then used to calibrate the poroelastic properties of the reservoir and caprock along with the geometry of an elliptic channel associated with the formation. Using our approach we are able to identify issues with model error as well as substantial tradeoffs that occur within the parameters.

inversion, stochastic, geomechanics, poroelastic, ↗

The [(AI 2 O 3 ) 2 ] - Anion Cluster: Electron Localization-Delocalization Isomerism

Three-dimensional bulk alumina and its two-dimensional thin films show great structural diversity, posing considerable challenges to their experimental structural characterization and computational modeling. Recently, structural diversity has also been demonstrated for zerodimensional gas phase aluminum oxide clusters. Mass-selected clusters not only make systematic studies of the structural and electronic properties as a function of size possible, but lately have also emerged as powerful molecular models of complex surfaces and solid catalysts. In particular, the [(Al 2 O 3 ) 3-5 ] + clusters were the first example of polynuclear maingroup metal oxide cluster that are able to thermally activate CH 4 . Over the past decades gas phase aluminum oxide clusters have been extensively studied both experimentally and computationally, but definitive structural assignments were made for only a handful of them: the planar [Al 3 O 3 ] - and [Al 5 O 4 ] - cluster anions, and the [(Al 2 O 3 ) 1-4 (AlO)] + cluster cations. For stoichiometric clusters only the atomic structures of [(Al 2 O 3 ) 4 ] +/0 have been nambiguously resolved. Here we report on the structures of the [(Al 2 O 3 ) 2 ] -/0 clusters combining photoelectron spectroscopy (PES) and quantum chemical calculations employing a genetic algorithm as a global optimization technique. The [(Al 2 O 3 ) 2 ] - cluster anion show energetically close lying but structurally distinct cage and sheet-like isomers which differ by delocalization/localization of the extra electron. The experimental results are crucial for benchmarking the different computational methods applied with respect to a proper description of electron localization and the relative energies for the isomers which will be of considerable value for future computational studies of aluminum oxide and related systems.

59 BASIC BIOLOGICAL SCIENCES↗