SEARCH · Search NASA
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.
Python scripts for physics informed machine learning for time-dependent problems. The code implements hard constrained sequential physics-informed neural networks (HCSPINNs) using JAX library. The details of the method and implementation can be found in the following paper: Roy, P., & Castonguay, S. (2024). Exact Enforcement of Temporal Continuity in Sequential Physics- Informed Neural Networks. arXiv preprint arXiv:2403.03223. (https://arxiv.org/abs/2403.03223)
This paper presents a 1-D convolutional and graph convolutional networks for fault detection in microgrids. The combination of 1-D convolutional neural networks (1D-CNN) and graph convolutional networks (GCN) helps extract both spatial-temporal correlations from the voltage measurements in microgrids. The fault detection scheme includes fault event detection, fault type and phase classification, and fault location. There are five neural network model training to handle these tasks. Transfer learning and fine-tuning are applied to reduce training efforts. The combined 1-D convolutional and graph convolutional networks (1D-CGCN) is compared with the traditional ANN structure on the Potsdam 13-bus microgrid dataset. The accuracy of 99.5%, 98.4%, 99.2%, and 95.5% are achieved in fault event detection, fault type classification, fault phase identification, and fault location respectively. The detailed confusion matrices of fault type and fault phase classification are provided for validation.
As the flexibility of generation and demand increases in distribution systems, the residential loads are emerging as a promising means to participate in demand response and the transactive energy market. Market pricing is an instrumental mechanism for the distribution system operator to exploit the full potential of the flexible resources. The distribution locational marginal price (DLMP) can be used to guide the residential load consumption. This type of market signal helps the distribution system operator to optimize the scheduling of all resources while satisfying related network constraints through a day-ahead market. However, solving the optimization problem for large-scale systems can be computationally expensive. To address the scalability and practicability limitations of the DLMP framework, a learning-based approach is proposed in this paper to complement the day-ahead distribution market framework. Here, the proposed approach combines long short-term memory and transfer learning to develop deep neural network that can capture the spatial–temporal correlation of the input data. The model can determine the optimal DLMP for each node in a distribution system without the system parameters required to formulate the optimization problem. Testing results on IEEE 33-bus and 123-bus systems show that the proposed approach can generate a comparable DLMP against the optimization solutions.
Not Available
The cost-effective and sustainable deployment of hydrogen supply and demand networks, especially in large economic regions like California, can be challenging considering the spatial-temporal availability and variability of the different actors across the network such as production processes, distribution modes, and end-users. In this presentation, we will provide an overview and demonstration of a modeling framework used to assess the environmental, economic, and human health impacts of plausible hydrogen production and supply chain networks in California. Scenarios focus on green hydrogen production pathways using water electrolysis and biomass gasification. End-use applications included in the model are transit, medium and heavy-duty trucking, port authorities, and power and aviation companies that currently consume natural gas, diesel, and aviation fuel for their day-to-day operation. Representative locations for hydrogen production and end-use are based on recent projections of the hydrogen economy in California. All mass and energy flows, as well as estimated emissions, are based on H2A process model designs and projections of technology performance, literature review, and LBNL process, economic and life cycle modeling, and not on company data for the sake of this presentation. Human health impacts are included following methodologies developed for the University of California Irvine HyDeal project. Life cycle phases associated with hydrogen production include feedstock preparation (water and biomass), energy production and consumption (renewable, grid, and combination of renewable and grid electricity), maintenance (chemical utilization in electrolysis and natural gas combustion in gasification), carbon sequestration, hydrogen storage (compression and liquefaction), and distribution (truck and pipeline). We apply the framework utilizing California specific emission factors, financial data, and human health damages and explore the impact of network characteristics on results. Example variations include: the inclusion of policy incentives or not, different representations of the electricity grid and source, electrolysis versus gasification versus combinations of both for production, liquefaction versus compression based on producer capacity cutoffs, transportation truck versus pipeline based on existing infrastructure, and ultimate end use. Comparison of these different scenarios can help inform future projects by demonstrating the trade-offs among environmental, economic, and human health impacts. This model, automated in R, is a starting platform upon which new analysis, modeling capabilities, locations, and emission factors can be rapidly tested and integrated.
Recurrent Neural Networks (RNN) is widely applied to temporal sequence analysis, where real-time performance is usually in demand. However, RNN suffers a heavy computational workload as the model comes with a large weight matrix. To alleviate the pain, model compression (pruning) schemes have been proposed for RNN that pruning the redundant (near-zero) weight-values. On the one hand, the non-structured pruning methods achieve a considerable pruning rate while bringing the computational irregularity, which is un-friendly to parallel-hardware. On the other hand, the existing structured pruning methods consider the hardware parallelism; However, they suffer a poor pruning rate due to the restrict constraints on pruning structure. This paper presents CSB-RNN, an optimized full-stack RNN framework with the novel compressed structured block (CSB) technique. The CSB-pruned RNN model comes with both fine-granularity that benefits the pruning rate and regular structure that facilitates the hardware-parallelism. Further, we propose a novel hardware architecture for inferencing the CSB-pruned model. Different from conventional parallel hardware, this architecture solves the block-workload imbalance issue and achieves an over 95% hardware utilization. With the experiments on 10 RNN models in 5 application domains, the CSB-RNN realizes 7×-20× lossless compression and up to 50× acceptable lossy-compression, which is 2×-7× to the prior art. With the addition of the novel hardware, the compressed-RNN inference reaches a super real-time latency of 10-400µs with FPGA implementation.
Operator learning for time-dependent partial differential equations (PDEs) has seen rapid progress in recent years, enabling efficient approximation of complex spatiotemporal dynamics. However, most existing methods rely on fixed time step sizes during rollout, which limits their ability to adapt to varying temporal complexity and often leads to error accumulation. In this work, we propose the Time-Adaptive Transformer with Neural Taylor Expansion (TANTE), a novel operator-learning framework that produces continuous-time predictions with adaptive step sizes. TANTE predicts future states by performing a Taylor expansion at the current state, where neural networks learn both the higher-order temporal derivatives and the local radius of convergence. This allows the model to dynamically adjust its rollout based on the local behavior of the solution, thereby reducing cumulative error and improving computational efficiency. We demonstrate the effectiveness of TANTE across a wide range of PDE benchmarks, achieving superior accuracy and adaptability compared to fixed-step baselines, delivering accuracy gains of 60-80 % and speed-ups of 30-40 % at inference time.
Meteorological prediction is crucial for various sectors, including agriculture, navigation, daily life, disaster prevention, and scientific research. However, traditional numerical weather prediction (NWP) models are constrained by their high computational resource requirements, while the accuracy of deep learning models remains suboptimal. In response to these challenges, we propose a novel deep learning-based model, the Spatiotemporal Fusion Model (STFM), designed to enhance the accuracy of meteorological predictions. Our model leverages Fifth-Generation ECMWF Reanalysis (ERA5) data and introduces two key components: a spatiotemporal encoder module and a spatiotemporal fusion module. The spatiotemporal encoder integrates the strengths of convolutional neural networks (CNNs) and recurrent neural networks (RNNs), effectively capturing both spatial and temporal dependencies. Meanwhile, the spatiotemporal fusion module employs a dual attention mechanism, decomposing spatial attention into global static attention and channel dynamic attention. This approach ensures comprehensive extraction of spatial features from meteorological data. The combination of these modules significantly improves prediction performance. Experimental results demonstrate that STFM excels in extracting spatiotemporal features from reanalysis data, yielding predictions that closely align with observed values. In comparative studies, STFM outperformed other models, achieving a 7% improvement in ground and high-altitude temperature predictions, a 5% enhancement in the prediction of the u/v components of 10 m wind speed, and an increase in the accuracy of potential height and relative humidity predictions by 3% and 1%, respectively. This enhanced performance highlights STFM’s potential to advance the accuracy and reliability of meteorological forecasting.
Artificial neural networks (ANNs) for performing spectroscopic gamma-ray source identification have been previously introduced, primarily for applications in controlled laboratory settings. To understand the utility of these methods in scenarios and environments more relevant to nuclear safety and security, this work examines the use of ANNs for mobile detection, which involves highly variable gamma-ray background, low signal-to-noise ratio measurements, and low false alarm rates. Simulated data from a 2” × 4” × 16” NaI(Tl) detector are used in this work for demonstrating these concepts, and the minimum detectable activity (MDA) is used as a performance metric in assessing model performance.In addition to examining simultaneous detection and identification, binary spectral anomaly detection using autoencoders is introduced in this work, and benchmarked using detection methods based on Non-negative Matrix Factorization (NMF) and Principal Component Analysis (PCA). On average, the autoencoder provides a 12% and 23% improvement over NMF- and PCA-based detection methods, respectively. Additionally, source identification using ANNs is extended to leverage temporal dynamics by means of recurrent neural networks, and these time-dependent models outperform their time-independent counterparts by 17% for the analysis examined here. The paper concludes with a discussion on tradeoffs between the ANN-based approaches and the benchmark methods examined here.
Top-down studies have found consistent underestimations in the United States Environmental Protection Agency (EPA) methane emissions inventory from the oil and gas (O&G) sector. Many of these studies use observations that bias toward hours when worktime activity occurs. In this study, we analyze over 2 years of methane measurements from a tower network in the Delaware basin to analyze hourly temporal emission patterns. Inversion results suggest a range in emissions from 137 Mg/hr at night to 197 Mg/hr during the day, present during both weekdays and weekends. If these results are applicable to other basins, daytime-influenced methodologies may overestimate daily emission rates by up to 27%. This bias does not reconcile the more than 200% difference between the EPA inventory and top-down estimates in the Delaware basin. This study demonstrates how continuous measurement networks can be combined with detailed activity data to improve bottom-up inventories.
When learning, the brain modifies individual synaptic connections to reach a desired behavior. Animal and human brains have been shown to be incredibly capable of learning complex and varied functions across a wide variety of tasks. In recent years, artificial neural networks, inspired by human and animal brains, have shown great capabilities in learning a wide variety of difficult tasks. However, artificial neural networks primarily teach themselves through the use of backpropagation, a learning method which has no clear analogue within the brain. Additionally, Artificial Neural Networks primarily use continuous activation functions, which differ significantly from the spiking neuronal behavior present in the brain. In this paper, we discuss and demonstrate a biologically plausible learning method that approximates backpropagation through two techniques on Spiking Neural Networks. First, we show that the local temporal derivatives that are necessary for backpropagation can be approximately recovered through reconstruction using spike timings. Second, we show that through learning during a sleep phase, inspired by neuroscience research into memory replay, the localized parallel feedback path can learn to approximate the derivative through the forward path weight matrix, thus solving the weight transport problem. Lastly, we demonstrate that the combination of these two methods can approach or exceed the accuracy of backpropagation-based methods for a variety of neuromorphic vision tasks while maintaining biological plausibility.
We apply an interpretable Long Short-Term Memory (iLSTM) network for land-atmosphere carbon flux predictions based on time series observations of seven environmental variables. iLSTM enables interpretability of variable importance and variable-wise temporal importance to the prediction of targets by exploring internal network structures. The application results indicate that iLSTM not only improves prediction performance by capturing different dynamics of individual variables, but also reasonably interprets the different contribution of each variable to the target and its different temporal relevance to the target. This variable and temporal importance interpretation of iLSTM advances terrestrial ecosystem model development as well as our predictive understanding of the system.
Abstract Profiles of atmospheric temperature and water vapor from remotely sensed platforms provide critical observations within the temporal and spatial gaps of the radiosonde network. The 2017 National Academies of Science Decadal Survey highlighted that observations of the planetary boundary layer (PBL) from the current space-based observing system are not of the necessary accuracy or resolution for monitoring and predicting high-impact weather phenomena. One possible solution to improving observations of the PBL is supplementing the existing space-based observing system with a network of ground-based profilers. A synthetic information content study is developed utilizing profiles from the Atmospheric Radiation Measurement (ARM) program sites at the Southern Great Plains (SGP), east North Atlantic (ENA), and North Slope of Alaska (NSA) to assess the benefits, in terms of degrees of freedom (DOF), vertical resolution, and uncertainties, of a synergy between the ground-based Atmospheric Emitted Radiance Interferometer (AERI) with space-based hyperspectral infrared (IR) sounders. A combination of AERI with any of the three polar-orbiting IR sounders: the Atmospheric Infrared Sounder (AIRS), the Cross-track Infrared Sounder (CrIS), or the Infrared Atmospheric Sounding Interferometer (IASI), results in a DOF increase of 30%–40% in the surface-to-700-hPa layer compared to the space-based instrument alone. Introducing AERI measurements to the observing system also results in significant improvements to vertical resolution and uncertainties in the bottom 1000 m of the atmosphere compared to CrIS measurements alone. A synergy of CrIS and AERI exceeds the 1-km-vertical-resolution goal set by the Decadal Survey in the lowest 1000 m.
In the face of the increasingly dire consequences of anthropogenic climate change, capturing and storing carbon dioxide is paramount. However, several impediments exist to the safe and effective subsurface storage of CO2, such as cost of transport, identification of suitable sites for subsurface storage, and assessment of long-term risk from storage in subsurface aquifers. Accurate subsurface modeling is necessary to ensure that CO2 storage is both safe and effective. Still, such modeling has traditionally required either substantial time and computational power (numerical simulation) or a substantial amount of pre-existing data for training (machine learning models). Additionally, these models lack flexibility in dealing with both changes in discretization of the input data and generalizability beyond the data on which they are trained. In order to address these issues, this research applies graph neural networks (GNNs) to predict subsurface saturation and pressure during CO₂ injection in a model of the Illinois Basin-Decatur Project (IBDP). GNNs provide a flexible, intuitive method for representing and manipulating complex unstructured data, which is often found in many practical domain problems such as fluid flow and subsurface characterization. These unstructured grids are easily represented in GNNs by representing spatially-localized features such as permeability, porosity, saturation, and pressure as nodes in a graph and relationships between these properties as edges connecting these nodes. This research applies a specific GNN model called MeshGraphNets (MGN) to model the change in CO2 saturation and pressure over a 50-month time period (36 months of injection, 14 months post-injection). The MGN model leverages a message passing process that allows the network to learn both the spatial and temporal dynamics of this system simultaneously. Additionally, training on a limited dataset (64 realizations, 20 time points each) resulted in a high degree of accuracy in saturation prediction both within the same timeframe as the training (20 months, 0.039 average RMSE) and when projecting out to the end of injection (36 months, 0.053 average RMSE). Temporal predictions such as those generated by MGNs and other similar models are prone to accumulated error over time; in order to address this, a multi-step rollout (MSR) training process was applied to calculate training loss. This method mimics the forward prediction during inference by “rolling out” multiple time points in a single training step using the previous prediction as input to the MGN model. By calculating the loss several time steps forward from the current prediction, the model is forced to find a more stable state over time. Application of MSR to the MGN model resulted in an average 15% reduction in inference error over time during forward prediction. This study showcases the immense potential of GNNs as a game-changing methodology for predicting pressure and saturation evolution in CCS projects, ultimately paving the way for more sustainable and effective carbon storage solutions. Presentation prepared for the 2024 AiChE Annual Meeting, October 27 to November 1 2024, San Diego, CA.
Over the past several years, many new data sets have become available regarding the characterization of the Utah FORGE reservoir. These include, but are not limited to, the stimulation of Well 16A, the drilling and completion of Well 16B, and interwell circulation confirmatory testing. As part of the characterization efforts, conceptual models of the reservoir are re-examined as new data become available. As part of the planning for FORGE activities, numerical models are often used to predict the reservoir response to the planned testing. Stochastic methods are often employed to bound uncertainty and allow for evaluation of comprehensive ranges of key reservoir parameters. For the most recent interwell circulation confirmatory testing (July 2023), a priori numerical model predictions did bound the observed behavior (Xinj et al., 2023), but key deviations from expected behavior prompted the FORGE team to reevaluate our conceptual model of the reservoir. In early October 2023, key members of the development, testing, and monitoring teams met for 2 days to review newly collected data and discuss ‘interesting’ or ‘key’ observations. From these discussions, 15 Key Observations were documented, with several significant ones being that the discrete fracture network developed from the 16A stimulation data may not be appropriate and that the early time pressure data obtained during the summer 2023 reservoir testing were best described using radial solutions. In July 2023, two campaigns of interwell confirmatory testing were conducted, the first set of tests occurred on July 4-5 and the second set on July 18-19. The second set of circulation tests conducted at the Utah FORGE site between the injection well 16A(78)-32 and production well 16B(78)-32 on July 18 and 19, 2023 are used to calibrate material properties in a thermal-hydraulic-mechanical (THM) simulation of the discrete fracture network connecting the wells. The spatially and temporally varying reservoir properties are calibrated to match the time dependent pressure and production profiles from the circulation tests. In future work, this calibrated model will be coupled to the native state THM model of the FORGE reservoir to predict surface deformation and strains resulting from pumping schedules.
Over the past several years, many new data sets have become available regarding the characterization of the Utah FORGE reservoir. These include, but are not limited to, the stimulation of Well 16A, the drilling and completion of Well 16B, and interwell circulation confirmatory testing. As part of the characterization efforts, conceptual models of the reservoir are re-examined as new data become available. As part of the planning for FORGE activities, numerical models are often used to predict the reservoir response to the planned testing. Stochastic methods are often employed to bound uncertainty and allow for evaluation of comprehensive ranges of key reservoir parameters. For the most recent interwell circulation confirmatory testing (July 2023), a priori numerical model predictions did bound the observed behavior (Xinj et al., 2023), but key deviations from expected behavior prompted the FORGE team to reevaluate our conceptual model of the reservoir. In early October 2023, key members of the development, testing, and monitoring teams met for 2 days to review newly collected data and discuss ‘interesting’ or ‘key’ observations. From these discussions, 15 Key Observations were documented, with several significant ones being that the discrete fracture network developed from the 16A stimulation data may not be appropriate and that the early time pressure data obtained during the summer 2023 reservoir testing were best described using radial solutions. In July 2023, two campaigns of interwell confirmatory testing were conducted, the first set of tests occurred on July 4-5 and the second set on July 18-19. The second set of circulation tests conducted at the Utah FORGE site between the injection well 16A(78)-32 and production well 16B(78)-32 on July 18 and 19, 2023 are used to calibrate material properties in a thermal-hydraulic-mechanical (THM) simulation of the discrete fracture network connecting the wells. The spatially and temporally varying reservoir properties are calibrated to match the time dependent pressure and production profiles from the circulation tests. In future work, this calibrated model will be coupled to the native state THM model of the FORGE reservoir to predict surface deformation and strains resulting from pumping schedules.
Biphoton frequency combs (BFCs) are promising quantum sources for large-scale and high-dimensional quantum information and networking systems. In this context, the spectral purity of individual frequency bins will be critical for realizing quantum networking protocols like teleportation and entanglement swapping. Measurement of the temporal autocorrelation function of the unheralded signal or idler photons comprising the BFC is a key tool for characterizing their spectral purity and in turn verifying the utility of the biphoton state for networking protocols. Yet the experimentally obtainable precision for measuring BFC correlation functions is often severely limited by detector jitter. The fine temporal features in the correlation function—not only of practical value in quantum information, but also of fundamental interest in the study of quantum optics—are lost as a result. We propose a scheme to circumvent this challenge through electro-optic phase modulation, experimentally demonstrating time-resolved Hanbury Brown–Twiss characterization of BFCs generated from an integrated 40.5-GHz Si 3 N 4 microring, up to a 3×3-dimensional two-qutrit Hilbert space. Through slight detuning of the electro-optic drive frequency from the comb’s free spectral range, our approach leverages Vernier principles to magnify temporal features, which would otherwise be averaged out by detector jitter. We demonstrate our approach under both continuous-wave and pulsed-pumping regimes, finding excellent agreement with theory. In conclusion, our method reveals not only the collective statistics of the contributing frequency bins but also their temporal shapes—features lost in standard fully integrated autocorrelation measurements.
Growing vaccine hesitancy is contributing to the decline in immunization rates for highly contagious, vaccine-preventable childhood diseases. Therefore, there has been a significant interest in understanding how hesitancy is spreading at higher spatio-temporal resolutions, enabling more targeted interventions. Motivated by this, we study the problem of prediction of vaccine hesitancy at the ZIP Code level, referred to as the VaxHesitancy problem. A significant challenge for this problem is the lack of high-resolution data that indicates hesitancy. Here, we develop a hybrid VaxHesSTL framework that combines a Graph Neural Network (GNN) and a Recurrent Neural Network (RNN) to address the VaxHesitancy problem. The GNN uses a ZIP Code-level network to capture spatial signals from neighboring areas, while the RNN models the temporal dynamics present in the data. We train and evaluate VaxHesSTL using a large dataset, namely the All-Payer Claims Databases (APCD), for Virginia, consisting of insurance claims from over five million individuals for six years. We find that an aggregated contact network or graph, developed from a detailed activity-based population network, plays an important role in the performance of VaxHesSTL, compared to graph models based solely on spatial proximity. Experiments demonstrate that VaxHesSTL outperforms a range of state-of-the-art baselines, which rely solely on historical time series data without accounting for spatial relationships. Since hesitancy data at higher spatial resolution is often unavailable or hard to get, we incorporate an active learning approach with our VaxHesSTL framework to optimize the training set without compromising the prediction performance. We find that hesitancy data for only 18% of ZIP Codes selected by active learning allows us to forecast hesitancy for all the ZIP Codes in the Virginia.