SEARCH · Engineering Papers

Results for “learning efficiency”

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.

SourceResearch connectionFrom yearThrough year

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

At least 811 records · Page 45

All sources NASA NTRS DOE OSTI

DOE OSTI2021

Inverse Aerodynamic Design of Gas Turbine Blades using Probabilistic Machine Learning

Abstract One of the critical components in Industrial Gas Turbines (IGT) is the turbine blade. Design of turbine blades needs to consider multiple aspects like aerodynamic efficiency, durability, safety and manufacturing, which make the design process sequential and iterative. The sequential nature of these iterations forces a long design cycle time, ranging from several months to years. Due to the reactionary nature of these iterations, little effort has been made to accumulate data in a manner that allows for deep exploration and understanding of the total design space. This is exemplified in the process of designing the individual components of the IGT resulting in a potential unrealized efficiency. To overcome the aforementioned challenges, we demonstrate a probabilistic inverse design machine learning framework, namely PMI (PMI), to carry out an explicit inverse design. PMI calculates the design explicitly without costly iteration and overcomes the challenges associated with ill-posed inverse problems. In this work the framework will be demonstrated on inverse aerodynamic design of three-dimensional turbine blades.

Engineering↗

DOE OSTI2021

Train Like a (Var)Pro: Efficient Training of Neural Networks with Variable Projection

Deep neural networks (DNNs) have achieved state-of-the-art performance across a variety of traditional machine learning tasks, e.g., speech recognition, image classification, and segmentation. The ability of DNNs to efficiently approximate high-dimensional functions has also motivated their use in scientific applications, e.g., to solve partial differential equations and to generate surrogate models. In this paper, we consider the supervised training of DNNs, which arises in many of the above applications. We focus on the central problem of optimizing the weights of the given DNN such that it accurately approximates the relation between observed input and target data. Devising effective solvers for this optimization problem is notoriously challenging due to the large number of weights, nonconvexity, data sparsity, and nontrivial choice of hyperparameters. To solve the optimization problem more efficiently, we propose the use of variable projection (VarPro), a method originally designed for separable nonlinear least-squares problems. Our main contribution is the Gauss--Newton VarPro method (GNvpro) that extends the reach of the VarPro idea to nonquadratic objective functions, most notably cross-entropy loss functions arising in classification. These extensions make GNvpro applicable to all training problems that involve a DNN whose last layer is an affine mapping, which is common in many state-of-the-art architectures. In our four numerical experiments from surrogate modeling, segmentation, and classification, GNvpro solves the optimization problem more efficiently than commonly used stochastic gradient descent (SGD) schemes. Finally, GNvpro finds solutions that generalize well, and in all but one example better than well-tuned SGD methods, to unseen data points.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2017

Energy-Efficient Homes (Done Right)

Home energy upgrade measures allow individuals to enjoy a safe, comfortable, and energy-efficient home. But how do you know if the work is done right? Learn how the U.S Department of Energy and its National Renewable Energy Laboratory are working to ensure that energy efficient measures are being installed correctly across the United States.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗

DOE OSTI2023

A Survey of Proposed Standards and Regulations for Artificial Intelligence

Over the last decade, the interest in machine learning and artificial intelligence (A.I) has grown exponentially. Researchers across several industries have explored and proposed uses of machine learning that have the potential to exponentially improve automation, efficiency, etc. The nuclear industry is no exception, with several studies exploring applications in almost every aspect of the field. However, these developments have brought forth a number of concerns. Government officials, industry leaders and many others have called for new regulations to govern the development and implementation of A.I systems. In order for machine learning and A.I to be viable for nuclear applications, there will need to be great regulatory oversight. The purpose of this study is to examine current and proposed regulations that relate to machine learning and provide recommendations for future regulations, specifically for nuclear applications.

99 GENERAL AND MISCELLANEOUS↗

DOE OSTI2026

Unsupervised atomic data mining via multi-kernel graph autoencoders for machine learning force fields

Constructing a chemically diverse dataset while avoiding sampling bias is critical to training efficient and generalizable force fields. However, in computational chemistry and materials science, many common dataset generation techniques are prone to oversampling regions of the potential energy surface. Furthermore, these regions can be difficult to identify and isolate from each other or may not align well with human intuition, making it challenging to systematically remove bias in the dataset. While traditional clustering and pruning (down-sampling) approaches can be useful for this, they can often lead to information loss or a failure to properly identify distinct regions of the potential energy surface due to difficulties associated with the high dimensionality of atomic descriptors. In this work, we introduce the Multi-kernel Edge Attention-based Graph Autoencoder (MEAGraph) model, an unsupervised approach for analyzing atomic datasets. MEAGraph combines multiple linear kernel transformations with attention-based message passing to capture geometric sensitivity and enable effective dataset pruning without relying on labels or extensive training. Demonstrated applications on niobium, tantalum, and iron datasets show that MEAGraph efficiently groups similar atomic environments, allowing for the use of basic pruning techniques for removing sampling bias. This approach provides an effective method for representation learning and clustering that can be used for data analysis, outlier detection, and dataset optimization.

Materials science↗

DOE OSTI2018

NREL Sustainable Mobility: The Transportation World Beyond Tomorrow – Is Here Today!

NREL’s sustainable mobility research is moving beyond vehicles and roads, investigating the best way to get people and goods where they need to go with maximum convenience, efficiency, and affordability while minimizing traffic congestion. Integrated research efforts span a variety of technologies—connected and autonomous vehicles and infrastructure; new materials and devices for top performance, efficiency, reliability, and durability; high-performance computing and machine learning; controls that integrate vehicles with energy, road, communication, building, and mass-transit systems; communities of the future, from rural to megacities; and consumer behavior and new models of car ownership.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

DOE OSTI2025

Quantum graph learning and algorithms applied in quantum computer sciences and image classification

Graph and network theory play a fundamental role in quantum computer sciences, including quantum information and computation. Random graphs and complex network theory are pivotal in predicting novel quantum phenomena, where entangled links are represented by edges. Quantum algorithms have been developed to enhance solutions for various network problems, giving rise to quantum graph computing and quantum graph learning (QGL). Here, in this review, we explore graph theory and graph learning methods as powerful tools for quantum computers to generate efficient solutions to problems beyond the reach of classical systems. We delve into the development of quantum complex network theory and its applications in quantum computation, materials discovery, and research. We also discuss quantum machine learning (QML) methodologies for effective image classification using qubits, quantum gates, and quantum circuits. Additionally, the paper addresses the challenges of QGL and algorithms, emphasizing the steps needed to develop flexible QGL solvers. This review presents a comprehensive overview of the fields of QGL and QML, highlights recent advancements, and identifies opportunities for future research.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

DOE OSTI2021

Quantum computing based hybrid deep learning for fault diagnosis in electrical power systems

Quantum computing (QC) and deep learning have shown promise of supporting transformative advances and have recently gained popularity in a wide range of areas. Here, this paper proposes a hybrid QC-based deep learning framework for fault diagnosis of electrical power systems that combine the feature extraction capabilities of conditional restricted Boltzmann machine with an efficient classification of deep networks. Computational challenges stemming from the complexities of such deep learning models are overcome by QC-based training methodologies that effectively leverage the complementary strengths of quantum assisted learning and classical training techniques. The proposed hybrid QC-based deep learning framework is tested on a simulated electrical power system with 30 buses and wide variations of substation and transmission line faults, to demonstrate the framework’s applicability, efficiency, and generalization capabilities. High computational efficiency is enjoyed by the proposed hybrid approach in terms of computational effort required and quality of diagnosis performance over classical training methods. In addition, superior and reliable fault diagnosis performance with faster response time is achieved over state-of-the-art pattern recognition methods based on artificial neural networks (ANN) and decision trees (DT).

42 ENGINEERING↗

DOE OSTI2014

TU-F-BRF-06: 3D Pancreas MRI Segmentation Using Dictionary Learning and Manifold Clustering

Purpose: The recent advent of MRI guided radiotherapy machines has lent an exciting platform for soft tissue target localization during treatment. However, tools to efficiently utilize MRI images for such purpose have not been developed. Specifically, to efficiently quantify the organ motion, we develop an automated segmentation method using dictionary learning and manifold clustering (DLMC). Methods: Fast 3D HASTE and VIBE MR images of 2 healthy volunteers and 3 patients were acquired. A bounding box was defined to include pancreas and surrounding normal organs including the liver, duodenum and stomach. The first slice of the MRI was used for dictionary learning based on mean-shift clustering and K-SVD sparse representation. Subsequent images were iteratively reconstructed until the error is less than a preset threshold. The preliminarily segmentation was subject to the constraints of manifold clustering. The segmentation results were compared with the mean shift merging (MSM), level set (LS) and manual segmentation methods. Results: DLMC resulted in consistently higher accuracy and robustness than comparing methods. Using manual contours as the ground truth, the mean Dices indices for all subjects are 0.54, 0.56 and 0.67 for MSM, LS and DLMC, respectively based on the HASTE image. The mean Dices indices are 0.70, 0.77 and 0.79 for the three methods based on VIBE images. DLMC is clearly more robust on the patients with the diseased pancreas while LS and MSM tend to over-segment the pancreas. DLMC also achieved higher sensitivity (0.80) and specificity (0.99) combining both imaging techniques. LS achieved equivalent sensitivity on VIBE images but was more computationally inefficient. Conclusion: We showed that pancreas and surrounding normal organs can be reliably segmented based on fast MRI using DLMC. This method will facilitate both planning volume definition and imaging guidance during treatment.

60 APPLIED LIFE SCIENCES↗

DOE OSTI2024

Efficiency of ML Anomaly Detection Triggers for Emerging Jets

Novel machine learning-based anomaly detection Level 1 (L1) triggers are currently under development at CMS, namely AXOL1TL and CICADA. The former employs a variational autoencoder, while the latter utilizes a convolutional autoencoder. These triggers aim to balance rate reduction with model independence, enabling the selection of potentially significant events that might be overlooked by traditional triggers relying on basic kinematic variable selections. Consequently, they have the potential to enhance signals indicative of physics beyond the Standard Model, such as those associated with emerging jets. Such signals are predicted by models featuring a composite dark sector where long-lived particles decay into Standard Model jets with displaced tracks and numerous vertices. This study evaluates the efficiency of these anomaly detection triggers in selecting events with emerging jets produced via the s-channel production of two dark quarks.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

DOE OSTI2024

Efficiency of ML Anomaly Detection Triggers for Emerging Jets

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

DOE OSTI2024

Efficiency of ML Anomaly Detection Triggers for Emerging Jets

43 PARTICLE ACCELERATORS↗

DOE OSTI2021

Bayesian sparse learning with preconditioned stochastic gradient MCMC and its applications

Deep neural networks have been successfully employed in an extensive variety of research areas, including solving partial differential equations. Despite its significant success, there are some challenges in effectively training DNN, such as avoiding overfitting in over-parameterized DNNs and accelerating the optimization in DNNs with pathological curvature. Here, we propose a Bayesian type sparse deep learning algorithm. The algorithm utilizes a set of spike-and-slab priors for the parameters in the deep neural network. The hierarchical Bayesian mixture will be trained using an adaptive empirical method. That is, one will alternatively sample from the posterior using preconditioned stochastic gradient Langevin Dynamics (PSGLD), and optimize the latent variables via stochastic approximation. The sparsity of the network is achieved while optimizing the hyperparameters with adaptive searching and penalizing. A popular SG-MCMC approach is Stochastic gradient Langevin dynamics (SGLD). However, considering the complex geometry in the model parameter space in nonconvex learning, updating parameters using a universal step size in each component as in SGLD may cause slow mixing. To address this issue, we apply a computationally manageable preconditioner in the updating rule, which provides a step-size parameter to adapt to local geometric properties. Moreover, by smoothly optimizing the hyperparameter in the preconditioning matrix, our proposed algorithm ensures a decreasing bias, which is introduced by ignoring the correction term in the preconditioned SGLD. According to the existing theoretical framework, we show that the proposed algorithm can asymptotically converge to the correct distribution with a controllable bias under mild conditions. Numerical tests are performed on both synthetic regression problems and learning solutions of elliptic PDE, which demonstrate the accuracy and efficiency of the present work.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

DOE OSTI2021

Data driven learning of robust nonlocal models

Nonlocal models use integral operators that embed length-scales in their definition. However, the integrands in these operators are difficult to define from the data that are typically available for a given physical system, such as laboratory mechanical property tests. In contrast, molecular dynamics (MD) does not require these integrands, but it suffers from computational limitations in the length and time scales it can address. To combine the strengths of both methods and to obtain a coarse-grained, homogenized continuum model that efficiently and accurately captures materials' behavior, we propose a learning framework to extract, from MD data, an optimal nonlocal model as a surrogate for MD displacements. Our framework guarantees that the resulting model is mathematically well-posed, physically consistent, and that it generalizes well to settings that are different from the ones used during training. The efficacy of this approach is demonstrated with several numerical tests for single layer graphene both in the case of perfect crystal and in the presence of thermal noise.

36 MATERIALS SCIENCE↗

DOE OSTI2025

Machine learning guided search for energetically favorable metal borocarbide ternary compounds

In this work, we employ machine-learning (ML) combined with first principles calculations to efficiently search for the energetically favorable metal borocarbide (M-B-C) ternary compounds with M being the group 1–3 metal elements. Using a crystal graph convolutional neural network (CGCNN) ML approach followed by first-principles calculations, we predicted 47 energetically favorable stable and metastable ternary Na-B-C, Ca-B-C, and La-B-C ternary compounds with their decomposition energy (E d ) below or within 100 meV/atom from the currently known convex hulls. Phonon spectra and electronic structures of the 14 energetically favorable stable structures are also investigated by first-principles calculations. By substituting the metal atoms in the 29 energetically favorable non-equivalent template structures of Na (Ca, La)-B-C with other group 1–3 elements in the periodic table, we further obtain 22 stable structures and 52 metastable structures (E d ≤100 meV/atom with respect to the known convex hulls) for Li-B-C, K-B-C, Rb-B-C, Mg-B-C, Sr-B-C, Ba-B-C, Sc-B-C and Y-B-C ternary compounds. New convex hulls including our newly predicted stable ternary structures and the known stable structures are constructed for the M-B-C systems. The results obtained by our ML guided first-principles calculations enrich our knowledge in the structure and energy landscape of metal borocarbide ternary compounds and provide useful guidance for further experimental synthesis and discovery.

36 MATERIALS SCIENCE↗

DOE OSTI2025

Physics-Informed Neural Network (PINN) Prediction of Mixed Mass-Heat-Crystallization Limited Methane Hydrate Formation and Dissociation in Micro-Confinement

The creation and use of Physics-Informed Neural Networks (PINNs) for simulating the dynamics of methane hydrate formation and dissociation will be presented. The PINN framework's main benefit is its capacity to impose physical consistency with only a partial comprehension of the governing equations. This makes the algorithm especially useful for systems with little experimental evidence or a lack of theoretical knowledge. A strong basis for forecasting methane hydrate behavior over the verified operating ranges of 30.0-80.9 bar pressure and 1.0-4.0 K sub-cooling conditions is provided by the combination of conductive heat transfer equations and mixed mass-transfer–crystallization kinetics. PINNs were more accurate at predicting the mixed mass-heat-crystallization limited kinetics than conventional Artificial Neural Networks (ANNs), demonstrating remarkable predictive accuracy for methane hydrate production over the ANN model. The efficiency of incorporating physical limitations from first principles into machine learning frameworks for methane hydrate crystallizations is reinforced by these findings. For hydrate-related applications in energy generation, carbon sequestration, and climate modelling, our study establishes PINNs as a computational tool that is both scalable and efficient. The proven capacity to close the gap between conventional physics-based simulations and solely data-driven models creates new opportunities for expedited hydrate research and practical applications.

Hartman, Ryan L [NYU Tandon School of Engineering]↗

DOE OSTI2025

Multi-head physics-informed neural networks for learning functional priors and uncertainty quantification

In numerous applications, the integration of prior knowledge and historical information is essential, particularly for tasks requiring the solution of ordinary or partial differential equations (ODEs/PDEs) in data-sparse or noisy environments. For instance, achieving accurate solutions to time-dependent PDEs with limited initial condition measurements necessitates an effective strategy for embedding prior knowledge. Hard-parameter sharing architectures in neural networks (NNs) have demonstrated success in both traditional and scientific machine learning domains, facilitating the learning of informative representations. Here, in this study, we introduce a novel, yet efficient, method to enhance physics-informed neural networks (PINNs) by incorporating a multi-head structure that enables the learning of functional priors from both empirical data and governing physical laws. This prior information can then be used to address data sparsity and high-level noise in solving ODE/PDE problems with uncertainty quantification (UQ). The approach, termed Multi-Head PINN (MH-PINN), consists of a shared body NN and multiple head NNs, each corresponding to an individual PINN instance. Our framework for functional prior learning is carried out in two stages: (1) training the MH-PINNs to develop a shared body NN alongside multiple head NNs, and (2) employing these trained head NNs to estimate a prior distribution through a normalizing flow-based density estimator. The learned functional prior can then be applied as a regularization mechanism in deterministic contexts or as an informative prior within a Bayesian inference framework, aiding in the resolution of subsequent ODE/PDE tasks. We evaluate the efficacy of MH-PINNs across five benchmark problems, including a high-dimensional parametric PDE, all characterized by data sparsity or substantial noise levels. Our findings reveal that MH-PINNs deliver accurate solutions and robust UQ, demonstrating adaptability across a range of complex and challenging scenarios.

Bayesian inference↗

DOE OSTI2015

Re-Defining Photovoltaic Efficiency Through Molecule Scale Control. Final Report

The direct conversion of sunlight into electricity represents one of the most important general means for sustainable energy production. Most modern photovoltaic cells are based on some form of semiconductor material such as silicon that is described by a characteristic energy or ''bandgap''. For solar rays with photon energy lower than the bandgap no absorption occurs and thus no energy is generated. For solar rays with photon energy greater than the bandgap, light may be absorbed to create a pair of electrical charges but only one bandgap of energy is created, leaving any additional energy lost as heat. Thus low bandgap materials can use a great part of the spectrum but can lose much of that energy; high bandgap materials fail to capture much of the spectrum. Thus there is a limit to the efficiency of such a solar cell that turns out to be about 32%. This limit is known as the Shockely-Queisser Limit. The Columbia EFRC program is dedicated to exploration of concepts that in principle can provide for efficiencies beyond this limit. One concept that this EFRC has explored for enhancing the efficiency of solar photovoltaic cells is called “Singlet Fission.” In this concept the absorption of light rays with photon energy at least twice the value of the basic bandgap for the system can produce two pairs of electrical charge carriers. If properly implemented this in principle can significantly reduce the energy lost as heat and give rise to solar cell efficiencies greater than the Shockley-Queisser limit. The problem is that there are virtually no materials that can undergo this process effectively. We have developed new materials that have demonstrated singlet fission efficiencies of almost 100%. We understand how these materials work and we have learned how to design many more systems in the future. So far we have only demonstrated the basic capability for efficient singlet fission. Much more work will be required to design and engineer specific materials that can be used practically in a solar cell system. In addition much work will be required to envision and demonstrate effective device structures that can utilize this concept. However these discoveries do provide the basis for an entirely new set of opportunities for more efficient solar energy generation moving beyond the Shockley-Queisser limit. A second part of the EFRC research program has been to investigate the material and device properties of an entirely new set of materials based on two-dimensional sheets (“ultra-thin”) with thicknesses of only one atom, or a single molecule or just a few atoms. These materials can exhibit conducting, insulating, and semiconducting character and thus they can form the basis for entirely new types of electrical devices. Recent fundamental investigations of these materials, at Columbia and elsewhere, demonstrate clearly that the flow of electrical charges in these systems is fundamentally different from the nature of electrical current flow in conventional materials. This fact presents many possibilities for new photovoltaic device concepts. The EFRC research team has achieved world leadership in the creation and understanding of these materials and in developing the fabrication techniques necessary to create useful devices from them. We have developed the basic fabrication methodology to build structures of these materials into complex device structures, layer by layer. Our EFRC research team has pioneered the synthesis and understanding for graphene, perhaps the simplest of these materials. Graphene can function as a highly transparent conducting material, capable of funneling an electrical charge over reasonable distances without significant energy loss. The EFRC program has also pioneered the development of ultra-thin sheets that function in a way analogous to semiconductor materials as well as sheets that act as electrical insulators. These developments therefore enable the construction of solar cells based on totally different physics from conventional cells. Because the electrons in these ultra-thin sheets interact strongly they will exhibit behaviors quite different from conventional materials with potential to operate at efficiencies beyond the Shockley-Queisser limit. In our EFRC program we have laid out many of the fundamental properties of these materials including the development of unique fabrication techniques. We discovered several new effects that demonstrate strong electron coupling. We have demonstrated the first solar cell that can actually generate electrical power at a high basic efficiency from these fundamentally new materials. There is much more science and technology required before these devices can become practical, but there is also very strong activity worldwide to build electronic devices from these materials, thus providing infrastructure and technical capability to develop these concepts.

14 SOLAR ENERGY↗