Results for “distributed learning” — Page 13

NASA NTRS1987

Learning in Artificial Neural Systems

This paper presents an overview and analysis of learning in Artificial Neural Systems (ANS's). It begins with a general introduction to neural networks and connectionist approaches to information processing. The basis for learning in ANS's is then described, and compared with classical Machine learning. While similar in some ways, ANS learning deviates from tradition in its dependence on the modification of individual weights to bring about changes in a knowledge representation distributed across connections in a network. This unique form of learning is analyzed from two aspects: the selection of an appropriate network architecture for representing the problem, and the choice of a suitable learning rule capable of reproducing the desired function within the given network. The various network architectures are classified, and then identified with explicit restrictions on the types of functions they are capable of representing. The learning rules, i.e., algorithms that specify how the network weights are modified, are similarly taxonomized, and where possible, the limitations inherent to specific classes of rules are outlined.

Matheus, Christopher J.↗

DOE OSTI2025

Visualization of Noisy and Less Noisy Computational Basis States in Quantum Computing

Quantum computing technology holds substantial promise as a reliable computational paradigm. However, current noisy intermediate scale quantum (NISQ) systems, are significantly impacted by noise originating from hardware inconsistencies. This noise causes errors and lowers output fidelity. So we must find which basis states cause errors. However, there are two main challenges in analyzing noise corresponding to basis states. First, the noise distribution data is high dimensional in nature, thereby making its analysis challenging. Second, although functional box plots have been used in the state of the art research to understand such a high dimensional data, they suffer from clutter and occlusion issues because of overplotting. In this study, we introduce an innovative visualization pipeline to address the aforementioned challenges to provide a clear depiction of noisy and less-noisy basis states. Specifically, our proposed visualization pipeline comprises three stages namely, low dimensional embedding, clustering, and violin plot visualization, to reduce visual clutter and effectively analyze high-dimensional noise distribution data. Our analysis uses quantum machine learning (QML) circuits as case study for drawing a distinction between noisy and less noisy basis states.

Senapati, Priyabrata [Kent State University]↗

DOE OSTI2025

Preliminary Results on Bayesian Inverse UQ for OECD/NEA WPNCS Subgroup 14 Benchmark Exercise for Error Recovery and Experimental Coverage

The Organization for Economic Cooperation and Development (OECD) Working Party on Nucelar Criticality Safety (WPNCS) has proposed a benchmark exercise representative of neutronic behavior in criticality experiments. Here, the goal is to develop confidence in data assimilation techniques used to adjust nuclear data. Participants are given synthetic experimental models with associated measured data and asked to estimate the model parameters given the model and measurements as well as provide predictions for separate application models. In this work, we performed data assimilation using Bayesian inverse Uncertainty Quantification (UQ) with machine learning surrogate models to produce posterior parameter distributions for the requested parameters and posterior predictive distributions for the requested responses. Several experimental models are shown to insufficiently inform the posterior parameter distributions for the applications involved. However, given sufficient experimental data, posterior parameter estimates yielded reduced uncertainty in the response predictions of interest while covering the experimental data.

Bayesian Inference↗

NASA NTRS1991

Experiments on neural network architectures for fuzzy logic

The use of fuzzy logic to model and manage uncertainty in a rule-based system places high computational demands on an inference engine. In an earlier paper, the authors introduced a trainable neural network structure for fuzzy logic. These networks can learn and extrapolate complex relationships between possibility distributions for the antecedents and consequents in the rules. Here, the power of these networks is further explored. The insensitivity of the output to noisy input distributions (which are likely if the clauses are generated from real data) is demonstrated as well as the ability of the networks to internalize multiple conjunctive clause and disjunctive clause rules. Since different rules with the same variables can be encoded in a single network, this approach to fuzzy logic inference provides a natural mechanism for rule conflict resolution.

Keller, James M.↗

NASA NTRS1993

MEDUSA - An overset grid flow solver for network-based parallel computer systems

Continuing improvement in processing speed has made it feasible to solve the Reynolds-Averaged Navier-Stokes equations for simple three-dimensional flows on advanced workstations. Combining multiple workstations into a network-based heterogeneous parallel computer allows the application of programming principles learned on MIMD (Multiple Instruction Multiple Data) distributed memory parallel computers to the solution of larger problems. An overset-grid flow solution code has been developed which uses a cluster of workstations as a network-based parallel computer. Inter-process communication is provided by the Parallel Virtual Machine (PVM) software. Solution speed equivalent to one-third of a Cray-YMP processor has been achieved from a cluster of nine commonly used engineering workstation processors. Load imbalance and communication overhead are the principal impediments to parallel efficiency in this application.

Smith, Merritt H.↗

NASA NTRS1994

OOD/OOP experience in the Science Operations Center part of the ground system for X ray Timing Explorer mission

The Science Operations Center (SOC) for the X-ray Timing Explorer (XTE) mission is an important component of the XTE ground system. Its mandate includes: (1) command and telemetry for the three XTE instruments, using CCSDS standards; (2) monitoring of the real-time science operations, reconfiguration of the experiment and the instruments, and real-time commanding to address the targets of opportunity (TOO) and alternate observations; and (3) analysis, processing, and archival of the XTE telemetry, and the timely delivery of the data products to the principal investigator (PI) teams and the guest observers (GO). The SOC has two major components: the science operations facility (SOF) that addresses the first two objectives stated above and the guest observer facility (GOF) that addresses the third. The SOF has subscribed to the object oriented design and implementation; while the GOF uses the traditional approach in order to take advantage of the existing software developed in support of previous missions. This paper details the SOF development using the object oriented design (OOD), and its implementation using the object oriented programming (OOP) in C++ under Unix environment on client-server architecture using Sun workstations. It also illustrates how the object oriented (OO) and the traditional approaches coexist in SOF and GOF, the lessons learned, and how the OOD facilitated the distributed software development collaboratively by four different teams. Details are presented for the SOF system, its major subsystems, its interfaces with the rest of the XTE ground data system, and its design and implementation approaches.

Choudhary, Abdur Rahim↗

NASA NTRS2005

Automated Planning and Scheduling for Space Mission Operations

Research Trends: a) Finite-capacity scheduling under more complex constraints and increased problem dimensionality (subcontracting, overtime, lot splitting, inventory, etc.) b) Integrated planning and scheduling. c) Mixed-initiative frameworks. d) Management of uncertainty (proactive and reactive). e) Autonomous agent architectures and distributed production management. e) Integration of machine learning capabilities. f) Wider scope of applications: 1) analysis of supplier/buyer protocols & tradeoffs; 2) integration of strategic & tactical decision-making; and 3) enterprise integration.

scheduling↗

NASA NTRSDate not supplied

Recent Progress on Rans-Based Transition Model Verification

The current efforts to assess and improve the Reynolds-averaged Navier-Stokes (RANS)-coupled transition models in the NASA FUN3D and OVERFLOW codes are summarized in this study. The first AIAA Transition Modeling Workshop and the NATO AVT-313 Transition Workshop both emphasized the need for code verification for transport equations based transition models as a top priority. We discuss the methods used for the model verification, the resulting grid families, the flow solutions, and other supporting information collected with at least two established NASA flow solvers, namely, FUN3D and OVERFLOW. These results, which will be uploaded onto the NASA Turbulence Modeling Resource, should assist other members of the computational fluid dynamics (CFD) community in verifying their own implementations of various transition models, such as the Langtry-Menter (LM2009) model, the one-equation γ model, and Coder’s amplification factor transport (AFT) model. Grid convergence is assessed using both global and local flow metrics of interest such as lift and drag as well as local skin-friction coefficients. We also explore the anisotropic unstructured metric-based adaptive mesh refinement library known as refine with the NASA FUN3D solver to determine if this capability can achieve the same accuracy as handcrafted structured grids with a significantly smaller node count and to learn the characteristics of the resulting grid distribution, especially in the vicinity of the transition zone.

Transition↗

NASA NTRSDate not supplied

Recent Progress on RANS-Based Transition Model Verification

The current efforts to assess and improve the Reynolds-averaged Navier-Stokes (RANS)-coupled transition models in the NASA FUN3D and OVERFLOW codes are summarized in this study. The first AIAA Transition Modeling Workshop and the NATO AVT-313 Transition Workshop both emphasized the need for code verification for transport equations based transition models as a top priority. We discuss the methods used for the model verification, the resulting grid families, the flow solutions, and other supporting information collected with at least two established NASA flow solvers, namely, FUN3D and OVERFLOW. These results, which will be uploaded onto the NASA Turbulence Modeling Resource, should assist other members of the computational fluid dynamics (CFD) community in verifying their own implementations of various transition models, such as the Langtry-Menter (LM2009) model, the one-equation γ model, and Coder’s amplification factor transport (AFT) model. Grid convergence is assessed using both global and local flow metrics of interest such as lift and drag as well as local skin-friction coefficients. We also explore the anisotropic unstructured metric-based adaptive mesh refinement library known as refine with the NASA FUN3D solver to determine if this capability can achieve the same accuracy as handcrafted structured grids with a significantly smaller node count and to learn the characteristics of the resulting grid distribution, especially in the vicinity of the transition zone.

RANS↗

DOE OSTI2024

Dynamically Learning Incentives for Load Control

As electrical generation becomes more distributed and volatile, and loads become more uncertain, controllability of distributed energy resources (DERs), regardless of their ownership status, will be necessary for grid reliability. Grid operators lack direct control over end-users' grid interactions, such as energy usage, but incentives can influence behavior -- for example, an end-user that receives a grid-driven incentive may adjust their consumption or expose relevant control variables in response. A key challenge in studying such incentives is the lack of data about human behavior, which usually motivates strong assumptions, such as distributional assumptions on compliance or rational utility-maximization. In this paper, we propose a general incentive mechanism in the form of a constrained optimization problem -- our approach is distinguished from prior work by modeling human behavior (e.g., reactions to an incentive) as an arbitrary unknown function. We propose feedback-based optimization algorithms to solve this problem that each leverage different amounts of information and/or measurements. We show that each converges to an asymptotically stable incentive with (near)-optimality guarantees given mild assumptions on the problem. Finally, we evaluate our proposed techniques in voltage regulation simulations on standard test beds. We test a variety of settings, including those that break assumptions required for theoretical convergence (e.g., convexity, smoothness) to capture realistic settings. In this evaluation, our proposed algorithms are able to find near-optimal incentives even when the reaction to an incentive is modeled by a theoretically difficult (yet realistic) function.

demand response↗

DOE OSTI2024

Voltage Mining for (De)lithiation-Stabilized Cathodes and a Machine Learning Model for Li-Ion Cathode Voltage

Advances in lithium-metal anodes have inspired interest in discovery of Li-free cathodes, most of which are natively found in their charged state. This is in contrast to today's commercial lithium-ion battery cathodes, which are more stable in their discharged state. In this study, we combine calculated cathode voltage information from both categories of cathode materials, covering 5577 and 2423 total unique structure pairs, respectively. The resulting voltage distributions with respect to the redox pairs and anion types for both classes of compounds emphasize design principles for high-voltage cathodes, which favor later Period 4 transition metals in their higher oxidation states and more electronegative anions like fluorine or polyanion groups. Generally, cathodes that are found in their charged, delithiated state are shown to exhibit voltages lower than those that are most stable in their lithiated state, in agreement with thermodynamic expectations. Deviations from this trend are found to originate from different anion distributions between redox pairs. In addition, a machine learning model for voltage prediction based on chemical formulas is trained and shows state-of-the-art performance when compared to two established composition-based ML models for material properties predictions, Roost and CrabNet.

25 ENERGY STORAGE↗

DOE OSTI2024

Federated Deep Reinforcement Learning for Decentralized VVO of BTM DERs

The future of grid control requires a hybrid approach combining centralized and decentralized methods to fully utilize the potential of smart edge devices with artificial intelligence (AI) capabilities. This paper aims to develop and evaluate a federated deep reinforcement learning (FDRL) framework for decentralized adaptive volt-var optimization (VVO) of behind-the-meter (BTM) distributed energy resources (DERs). First, this paper models a single deep reinforcement learning (DRL) agent using the Markov Decision Process (MDP) framework for decentralized adaptive VVO of BTM DERs. Two DRL algorithms, soft actor-critic (SAC) and twin-delayed deep deterministic policy gradient (TD3), are compared for their effectiveness in optimizing VVO. Results show that TD3 outperforms SAC, achieving a 71.3% improvement in mean reward. Finally, the DRL agent is deployed within the FDRL framework, using the Flower platform, to enhance learning, provide adaptive control, and ensure data privacy for BTM DERs.

Ravi, Abhijith↗

DOE OSTI2025

Contrastive learning for robust representations of neutrino data

In neutrino physics, analyses often depend on large simulated datasets, making it essential for models to generalize effectively to real-world detector data. Contrastive learning, a well-established technique in deep learning, offers a promising solution to this challenge. By applying controlled data augmentations to simulated data, contrastive learning enables the extraction of robust and transferable features. This improves the ability of models trained on simulations to adapt to real experimental data distributions. In this paper, we investigate the application of contrastive learning methods in the context of neutrino physics. Through a combination of empirical evaluations and theoretical insights, we demonstrate how contrastive learning enhances model performance and adaptability. Additionally, we compare it to other domain adaptation techniques, highlighting the unique advantages of contrastive learning for this field. Published by the American Physical Society 2025

Wilkinson, Alex (ORCID:0000000253404506)↗

DOE OSTI2025

A Dynamic Pricing Method to Manage the Impact of EV Charging on the Grid Using RL

This work addresses the challenge of managing electrical vehicle (EV) charging loads on distribution feeders with the increase in deployment of fast charging stations. To mitigate the adverse impacts on feeder health, a novel dynamic grid-informed pricing approach is proposed. This approach leverages reinforcement learning (RL) to determine hourly charging prices based on real-time grid conditions. A synthetic environment was developed to train the reinforcement learning agent. A model of an IEEE 34-bus distribution feeder with EV charging stations has been developed in OpenDSS utilizing Caldera for realistic EV charging profiles. Test cases demonstrate that the dynamic pricing strategy achieves higher energy delivery to the EV end user at a lower cost compared to constant pricing methods, while lowering voltage deviations and congestion. This approach offers more granular price adjustments, responding dynamically to feeder conditions and potentially improving grid stability and efficiency. The communication architecture to implement this dynamic pricing method is described. This research contributes to the development of smart grid-informed charging solutions that can reduce the cost of charging to the end user while also helping the grid.

EV charging, dynamic pricing, grid-informed chargi↗

DOE OSTI2025

A dynamic pricing method to manage the impact of EV charging on the grid using RL

This work addresses the challenge of managing electrical vehicle (EV) charging loads on distribution feeders with the increase in deployment of fast charging stations. To mitigate the adverse impacts on feeder health, a novel dynamic grid-informed pricing approach is proposed. This approach leverages reinforcement learning (RL) to determine hourly charging prices based on real-time grid conditions. A synthetic environment was developed to train the reinforcement learning agent. A model of an IEEE 34-bus distribution feeder with EV charging stations has been developed in OpenDSS utilizing Caldera for realistic EV charging profiles. Test cases demonstrate that the dynamic pricing strategy achieves higher energy delivery to the EV end user at a lower cost compared to constant pricing methods, while lowering voltage deviations and congestion. This approach offers more granular price adjustments, responding dynamically to feeder conditions and potentially improving grid stability and efficiency. The communication architecture to implement this dynamic pricing method is described. This research contributes to the development of smart grid-informed charging solutions that can reduce the cost of charging to the end user while also helping the grid.

24 - POWER TRANSMISSION AND DISTRIBUTION↗

DOE OSTI2026

Benchtop Autonomous Electrochemical Characterization System for Combinatorial Thin-Film Solid Oxide Electrodes

The design of materials for electrochemical energy conversion is complicated by a vast search space of candidate materials and multifaceted property requirements: multicarrier conductivity, stability, and catalytic activity are all necessary but rarely intersect. Although self-driving laboratories are rapidly rising to address such material optimization problems, the required infrastructure for integrated, large-scale robotic facilities can be cost-prohibitive. Here we develop and evaluate a closed-loop measurement system for efficient screening of proton-conducting oxide electrodes for ceramic fuel cells and electrolyzers, building on top of an existing benchtop instrument and integrating techniques for rapid impedance measurement and automated analysis. This system exemplifies a “minimum viable” self-driving implementation that can deliver substantial benefits with relatively simple infrastructure. Combinatorial thin-film microelectrode libraries are characterized with a recently developed joint time-domain and frequency-domain impedance measurement technique, which provides an order-of-magnitude acceleration relative to conventional impedance spectroscopy. The distribution of relaxation times is extracted from impedance data and analyzed without human intervention. These results feed an active learning and Bayesian optimization process that learns to predict electrochemical impedance as a function of material composition, measurement temperature, oxygen partial pressure, and electrical bias, which further reduces the screening time by tenfold with optimized experimental sequences. We apply this system to Ba⁡(Co,Fe,Zr,Y)⁢O 3−𝛿 combinatorial libraries and evaluate its effectiveness for learning material property trends and optimizing expensive-to-evaluate properties such as activation energy. This offers insights into key methodological aspects of practical autonomous experimentation, including surrogate model validation, cost-aware acquisition functions, and high-throughput data interpretation. Our results demonstrate the efficacy of the system for rapidly gathering information, but also highlight real-world experimental challenges of thin-film degradation and numerical instability in surrogate models.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

NASA NTRS1996

The cerebellum: a neuronal learning machine?

Comparison of two seemingly quite different behaviors yields a surprisingly consistent picture of the role of the cerebellum in motor learning. Behavioral and physiological data about classical conditioning of the eyelid response and motor learning in the vestibulo-ocular reflex suggests that (i) plasticity is distributed between the cerebellar cortex and the deep cerebellar nuclei; (ii) the cerebellar cortex plays a special role in learning the timing of movement; and (iii) the cerebellar cortex guides learning in the deep nuclei, which may allow learning to be transferred from the cortex to the deep nuclei. Because many of the similarities in the data from the two systems typify general features of cerebellar organization, the cerebellar mechanisms of learning in these two systems may represent principles that apply to many motor systems.

Non-NASA Center↗

NASA NTRS2017

NASA Computational Case Study: Spectral Energy Distribution Fitting

The need for faster, more efficient algorithms is an important aspect of scientific computing. Generally, scientists are only exposed to computational issues that arise in their field. Thus, collaboration between a numerical analyst and a scientist is becoming necessary for scientific computing. The purpose of this case study is to expose computer scientists to processes that an astronomer would use to obtain useful results from raw data. For example, astronomers are interested in determining the properties of galaxies and measuring changes in those properties as a function of time throughout cosmic history. To do so, they use certain models that are designed and refined over time via observations at different wavelengths of the light spectrum. The process of matching these models with observed data from studying celestial bodies is referred to as Spectral Energy Distribution (SED) fitting. In this case study, we learn how to perform the SED fit. This process requires knowledge of both the astronomical and computational issues involved when fitting flux, the total energy from a source as seen from Earth, to a set of physical templates. Once a fit is complete, one can classify the source and estimate a number of physical parameters. The goal is to demonstrate how the computer science skill set can be used in the scientific community and to possibly improve one or more of the computational aspects of this problem. The following provides some background on the astronomy issues, including information on the spectral energy distribution and related physical parameters, and the computational issues, including the fitting procedure..

modeling↗