Results for “learning dynamics” — Page 18

DOE OSTI2026

Data-driven Mori–Zwanzig modeling of Lagrangian particle dynamics in turbulent flows

The dynamics of Lagrangian particles in turbulence play a crucial role in mixing, transport, and dispersion in complex flows. Their trajectories exhibit highly nontrivial statistical behavior, motivating the development of surrogate models that can reproduce these trajectories without incurring the high computational cost of direct numerical simulations of the full Eulerian field. This task is particularly challenging because reduced-order models typically lack access to the full set of interactions with the underlying turbulent field. Novel data-driven machine learning techniques can be powerful in capturing and reproducing complex statistics of the reduced-order/surrogate dynamics. In this work, we show how one can learn a surrogate dynamical system that is able to evolve a turbulent Lagrangian trajectory in a way that is point-wise accurate for short-time predictions (with respect to Kolmogorov time) and stable and statistically accurate at long times. This approach is based on the Mori–Zwanzig formalism, which prescribes a mathematical decomposition of the full dynamical system into resolved dynamics that depend on the current state and the past history of a reduced set of observables, and the unresolved orthogonal dynamics due to unresolved degrees of freedom of the initial state. We show how by training this reduced order model on a point-wise error metric on short time-prediction, we are able to correctly learn the dynamics of Lagrangian turbulence, such that also the long-time statistical behavior is stably recovered at test time. This opens up a range of applications, for example, for the control of active Lagrangian agents in turbulence.

97 MATHEMATICS AND COMPUTING↗

NASA NTRS1992

Dynamic Optimization

We distinguish static and dynamic optimization of programs: whereas static optimization modifies a program before runtime and is based only on its syntactical structure, dynamic optimization is based on the statistical properties of the input source and examples of program execution. Explanation-based generalization is a commonly used dynamic optimization method, but its effectiveness as a speedup-learning method is limited, in part because it fails to separate the learning process from the program transformation process. This paper describes a dynamic optimization technique called a learn-optimize cycle that first uses a learning element to uncover predictable patterns in the program execution and then uses an optimization algorithm to map these patterns into beneficial transformations. The technique has been used successfully for dynamic optimization of pure Prolog.

Laird, Philip↗

NASA NTRS2005

Dynamic testing and simulation of the Mars Exploration Rover

In January of 2004, NASA landed two mobile robotic spacecraft on the surface of Mars as part of the Mars Exploration Rover (MER) project. The results and lessons learned of the test and dynamic simulation of the MER vehicles is provided in this paper.

rovers↗

NASA NTRS1992

A composite self tuning strategy for fuzzy control of dynamic systems

The feature of self learning makes fuzzy logic controllers attractive in control applications. This paper proposes a strategy to tune the fuzzy logic controller on-line by tuning the data base as well as the rule base. The structure of the controller is outlined and preliminary results are presented using simulation studies.

Shieh, C.-Y.↗

DOE OSTI2026

Continual Learning for Particle Accelerators

Particle accelerators operate under dynamically changing conditions, which often lead to data distribution drifts. These drifts pose significant challenges for Machine Learning (ML) models, which typically fail to maintain performance when faced with such non-stationary data. In particle accelerators, the primary sources of these data drifts include changes in accelerator settings and non-measured parameters such as machine degradation and environmental factors. Previous research has proposed conditional models to handle multiple beam configurations effectively; however, it is challenging to train the ML models on all possible configuration settings. Additionally, conditional models alone can not address performance degradation caused by drifts due to non-measured factors. These limitations contribute to a significant gap between ML development and its deployment in real-world operational settings. To bridge this gap, in this paper, we identify some of the key areas within particle accelerators where continual learning can help mitigate drift-induced performance degradation. In addition, we present a practical use case where a conditional Auto-Encoder model coupled with memory-based continual learning has been employed to demonstrate stable performance even when underlying data drifts.

Schram, Malachi [Thomas Jefferson National Acceler↗

DOE OSTI2026

Continual Learning for Particle Accelerators

Particle accelerators operate under dynamically changing conditions, which often lead to data distribution drifts. These drifts pose significant challenges for Machine Learning (ML) models, which typically fail to maintain performance when faced with such non-stationary data. In particle accelerators, the primary sources of these data drifts include changes in accelerator settings and non-measured parameters such as machine degradation and environmental factors. Previous research has proposed conditional models to handle multiple beam configurations effectively; however, it is challenging to train the ML models on all possible configuration settings. Additionally, conditional models alone can not address performance degradation caused by drifts due to non-measured factors. These limitations contribute to a significant gap between ML development and its deployment in real-world operational settings. To bridge this gap, in this paper, we identify some of the key areas within particle accelerators where continual learning can help mitigate drift-induced performance degradation. In addition, we present a practical use case where a conditional Auto-Encoder model coupled with memory-based continual learning has been employed to demonstrate stable performance even when underlying data drifts

Rajput, Kishansingh [Thomas Jefferson National Acc↗

DOE OSTI2026

Variational Optical Phase Learning on a Continuous-Variable Quantum Compiler

Quantum process learning is a fundamental primitive that draws inspiration from machine learning with the goal of better studying the dynamics of quantum systems. One approach to quantum process learning is quantum compilation, whereby an analog quantum operation is digitized by compiling it into a series of basic gates. While there has been significant focus on quantum compiling for discrete-variable systems, the continuous-variable (CV) framework has received comparatively less attention. We present an experimental implementation of a CV quantum compiler that uses two-mode squeezed light to learn a Gaussian unitary operation. We demonstrate the compiler by learning a parameterized linear phase unitary through the use of target and control phase unitaries to demonstrate a factor of 5.4 increase in the precision of the phase estimation and a 3.6-fold acceleration in the time-to-solution metric when leveraging quantum resources. We further show how our approach can be extended to higher-dimensional compilation tasks. Our results are enabled by the tunable control of our cost landscape via variable squeezing, thus providing a critical framework to simultaneously increase precision and reduce time-to-solution.

97 MATHEMATICS AND COMPUTING↗

NASA NTRS2009

Methodologies for Adaptive Flight Envelope Estimation and Protection

This paper reports the latest development of several techniques for adaptive flight envelope estimation and protection system for aircraft under damage upset conditions. Through the integration of advanced fault detection algorithms, real-time system identification of the damage/faulted aircraft and flight envelop estimation, real-time decision support can be executed autonomously for improving damage tolerance and flight recoverability. Particularly, a bank of adaptive nonlinear fault detection and isolation estimators were developed for flight control actuator faults; a real-time system identification method was developed for assessing the dynamics and performance limitation of impaired aircraft; online learning neural networks were used to approximate selected aircraft dynamics which were then inverted to estimate command margins. As off-line training of network weights is not required, the method has the advantage of adapting to varying flight conditions and different vehicle configurations. The key benefit of the envelope estimation and protection system is that it allows the aircraft to fly close to its limit boundary by constantly updating the controller command limits during flight. The developed techniques were demonstrated on NASA s Generic Transport Model (GTM) simulation environments with simulated actuator faults. Simulation results and remarks on future work are presented.

Tang, Liang↗

DOE OSTI2026

High-Performance Semiempirical Excited-State Molecular Dynamics Powered by Graphics Processing Units

Here, this Letter introduces excited-state molecular dynamics in PYSEQM, a GPU-accelerated semiempirical quantum chemistry engine implemented in PyTorch. The new module enables Born–Oppenheimer molecular dynamics (BOMD) using configuration-interaction singles and random phase approximation for excited states, allowing long trajectories and large statistical ensembles to be simulated efficiently on a single GPU. We also implement an extended Lagrangian excited-state BOMD (XL-ESMD) scheme that propagates auxiliary electronic variables, enabling relaxed ground and excited-state convergence thresholds without compromising energy conservation. The excited-state BOMD implementation scales smoothly from small chromophores to a nearly 900-atom dendrimer (taking 6.5 s per MD step). PYSEQM also supports batched execution, allowing many geometries or trajectories to be evaluated in a single GPU launch, substantially increasing throughput and making ensemble-based protocols routine. As a demonstration, we compute absorption, emission, and infrared spectra from trajectories propagated on the ground and first excited states. The XL-ESMD scheme yields identical spectra at significantly lower computational cost, establishing the role of extended Lagrangian based dynamics for efficient excited-state BOMD simulations. Beyond raw performance, PYSEQM’s PyTorch foundation provides automatic differentiation for forces, efficient GPU batching, and seamless interfacing with machine learning models. These capabilities position PYSEQM as a practical platform for machine learning-augmented excited-state dynamics and lay the foundation for future data-driven nonadiabatic excited-state dynamics modeling of ultrafast spectroscopic probes.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

DOE OSTI2025

Spatio-temporal Fourier Transformer for Long-term Dynamics Prediction (StFT) v1.0

We propose a novel machine learning model spatio-temporal Fourier transformer (StFT) to emulate long-term dynamics of multi-scale and multi-physics systems. Our method StFT overcomes the limitations of rapid error accumulation, particularly in long-term forecasting of systems characterized by complex and coupled dynamics. StFT achieves outstanding accuracy and computational efficiency by effectively capturing multi-scale interactions, and quantify the uncertainties inherent in the predictions. Our model leverages a structured hierarchy of StFT blocks, and explicitly captures dynamics across both macro- and micro- spatial scales. Evaluations conducted on three benchmark datasets (plasma, fluid, and atmospheric dynamics) demonstrate the advantages of our approach over state-of-the-art ML methods.

Bai, Zhe [Lawrence Berkeley National Laboratory (L↗

DOE OSTI2025

Phonon second harmonic generation in NaBr studied by inelastic neutron scattering and computer simulation

The phenomenon of second harmonic generation (SHG) was found for phonons in anharmonic NaBr by inelastic neutron scattering. The temperature dependence of this phonon SHG was measured from 300 K to 650 K. At 300 K the second harmonic (SH) is seen as a high-energy branch around 33 meV, nearly independent of $\overrightarrow{Q}$. The temperature effective potential (TDEP) method and classical molecular dynamics (MD) simulation with machine learning interatomic potential were able to reproduce the SH, and showed that SHG occurs with the flat transverse optical (TO) phonon branch. A classical model of a nonlinear medium explains the intensity and lifetime of the SH, compared to those of the TO modes. Also successful was a quantum model based on the Heisenberg-Langevin equation for interacting phonons coupled to a thermal bath, which also predicts a spectral distribution of the SH. In conclusion, the measured temperature dependence of the intensity of the second harmonic showed that it follows the Planck distribution of a one-phonon quasiparticle, and not two TO phonons.

36 MATERIALS SCIENCE↗

NASA NTRS2003

Intelligent Launch and Range Operations Virtual Test Bed (ILRO-VTB)

Intelligent Launch and Range Operations Virtual Test Bed (ILRO-VTB) is a real-time web-based command and control, communication, and intelligent simulation environment of ground-vehicle, launch and range operation activities. ILRO-VTB consists of a variety of simulation models combined with commercial and indigenous software developments (NASA Ames). It creates a hybrid software/hardware environment suitable for testing various integrated control system components of launch and range. The dynamic interactions of the integrated simulated control systems are not well understood. Insight into such systems can only be achieved through simulation/emulation. For that reason, NASA has established a VTB where we can learn the actual control and dynamics of designs for future space programs, including testing and performance evaluation. The current implementation of the VTB simulates the operations of a sub-orbital vehicle of mission, control, ground-vehicle engineering, launch and range operations. The present development of the test bed simulates the operations of Space Shuttle Vehicle (SSV) at NASA Kennedy Space Center. The test bed supports a wide variety of shuttle missions with ancillary modeling capabilities like weather forecasting, lightning tracker, toxic gas dispersion model, debris dispersion model, telemetry, trajectory modeling, ground operations, payload models and etc. To achieve the simulations, all models are linked using Common Object Request Broker Architecture (CORBA). The test bed provides opportunities for government, universities, researchers and industries to do a real time of shuttle launch in cyber space.

Bardina, Jorge↗

NASA NTRSDate not supplied

Predicting Melt Properties Using Atomistic Simulations With A Highly Accurate Physically Informed Neural Network Interatomic Potential

The use of a recently developed machine learning (ML) interatomic potential for molecular dynamics simulations of aluminum melt properties will be presented. Such properties are critical for process modeling in additive manufacturing, including the melt pool size, solidification, and formation of solidification microstructures. Direct first-principles modeling of these processes is computationally prohibitive whereas simulations employing ML potentials combine the high accuracy of quantum-mechanical methods with high computational speeds. The physically-informed neural network (PINN) method used herein, integrates a high-dimensional regression implemented by an artificial neural network with a physics-based bond-order interatomic potential. PINN potentials can accurately reproduce many properties of aluminum in both crystalline-solid and liquid phases. We examine the accuracy of a PINN Al potential in predicting the density, self-diffusivity, viscosity, and the tension of the liquid surface and liquid-solid interfaces. Comparison with experimental data and ab initio molecular dynamics calculations shows very good agreement for all properties tested.

molecular dynamics↗

DOE OSTI2026

Reduced‐Order Modeling of Energetic Materials Using Physics‐Aware Recurrent Convolutional Neural Networks in a Latent Space (LatentPARC)

Physics-aware deep learning (PADL) has gained popularity for use in spatiotemporal dynamics simulations, such as those in computational modeling of energetic materials (EM). We show that the challenge PADL methods face while learning complex field evolution problems can be simplified and accelerated by decoupling it into two tasks: learning complex geometric features in evolving fields and modeling dynamics over these features in a lower-dimensional feature space. We build upon our previous work on physics-aware recurrent convolutional neural networks (PARC). PARC embeds knowledge of underlying physics into its neural network architecture for more robust and accurate prediction of evolving physical fields. PARC was shown to effectively learn complex nonlinear features such as the formation of hotspots and coupled shock fronts in various initiation scenarios of EMs, as a function of microstructures, serving effectively as a microstructure-aware burn model. Here, we further accelerate PARC and reduce its computational cost by projecting the original dynamics onto a lower-dimensional invariant manifold, or “latent space.” The projected latent representation encodes the complex geometry of evolving fields (e.g., temperature and pressure) in a set of data-driven features. The reduced dimension of this latent space allows us to learn the dynamics during the initiation of EM with a lighter and more efficient model. We observe a significant decrease in training and inference time while maintaining results comparable to PARC at inference. This work takes steps towards enabling rapid prediction of EM thermomechanics at larger scales and characterization of EM structure–property–performance linkages at a full application scale.

Mathematics and Computing↗

DOE OSTI2024

Short-Term Probabilistic Solar Forecasting via Reinforcement Learning over ECMWF

In this paper, we present an innovative reinforcement learning approach for short-term solar forecasting, leveraging data from the European Centre for Medium-Range Weather Forecasts (ECMWF). The methodology begins with the application of the System Advisor Model (SAM) to transform various ECMWF numerical weather prediction members into predictive photovoltaic power generation. To enhance the precision of deterministic forecasting, we introduce a dynamic model selection algorithm based on Q-learning. This algorithm dynamically identifies and utilizes the most accurate ensemble member for forecasting purposes. Furthermore, we employ a support vector regression surrogate model with a Gaussian distribution to generate probabilistic forecasts, providing a holistic view of solar energy generation uncertainty. To expedite the training process and make it more practical for real-world applications, we integrate a rolling update workflow. This innovative workflow reduces the training period from months to a mere 19 days, making our method highly efficient. Numerical results of the case study show that in comparison to benchmark models, the proposed method improves the deterministic and probabilistic solar forecasting accuracy by up to 40.84% and 48.42%, respectively.

ensemble forecasting↗

NASA NTRSDate not supplied

Dynamic Spectrum Allocation in Urban Air Transportation System via Deep Reinforcement Learning

The emerging concepts of Urban Air Mobility (UAM) and Advanced Air Mobility (AAM) open a new paradigm for urban air transportation. A big challenge is that these new aerial vehicles will quickly saturate the already crowded aviation spectrum, which is an essential resource to ensure reliable communications for safe operations. In this paper, we consider an air transportation system where multiple aerial vehicles are operated to transport passengers or cargo from different sources to destinations along their pre-defined paths. During the flight, the minimum communication Quality of Service (QoS) requirement must be achieved to ensure flight safety. Our objective is to minimize the average mission completion time by jointly optimizing the velocity selection and spectrum allocation for all aerial vehicles. We formulate the optimization problem as a multi-stage Markov Decision Process (MDP) where the optimization variables are coupled together. A multi-agent Deep Reinforcement Learning (DRL) based solution is proposed where Value Decomposition Networks (VDN) algorithm is utilized to take discrete actions. Additionally, we propose a heuristic greedy algorithm as a baseline solution. Simulation results show that our learning based solution outperforms the heuristic greedy algorithm and another Orthogonal Multiple Access (OMA) solution in minimizing the mission completion time.

Ruixuan Han↗

NASA NTRSDate not supplied

Dynamic Spectrum Allocation in Urban Air Transportation System via Deep Reinforcement Learning

The emerging concepts of Urban Air Mobility (UAM) and Advanced Air Mobility (AAM) open a new paradigm for urban air transportation. A big challenge is that these new aerial vehicles will quickly saturate the already crowded aviation spectrum, which is an essential resource to ensure reliable communications for safe operations. In this paper, we consider an air transportation system where multiple aerial vehicles are operated to transport passengers or cargo from different sources to destinations along their pre-defined paths. During the flight, the minimum communication Quality of Service (QoS) requirement must be achieved to ensure flight safety. Our objective is to minimize the average mission completion time by jointly optimizing the velocity selection and spectrum allocation for all aerial vehicles. We formulate the optimization problem as a multi-stage Markov Decision Process (MDP) where the optimization variables are coupled together. A multi-agent Deep Reinforcement Learning (DRL) based solution is proposed where Value Decomposition Networks (VDN) algorithm is utilized to take discrete actions. Additionally, we propose a heuristic greedy algorithm as a baseline solution. Simulation results show that our learning based solution outperforms the heuristic greedy algorithm and another Orthogonal Multiple Access (OMA) solution in minimizing the mission completion time.

Ruixuan Han↗

DOE OSTI2026

Enabling probabilistic learning on manifolds through double diffusion maps

Here, we present a generative learning framework for probabilistic sampling that extends Probabilistic Learning on Manifolds (PLoM), which is designed to generate statistically consistent realizations of a random vector in a finite-dimensional Euclidean space, informed by a (representative) set of observations. In its original form, PLoM constructs a reduced-order probabilistic model by combining three main components: (a) kernel density estimation to approximate the underlying probability measure, (b) Diffusion Maps to characterize the manifold of the data, and (c) a reduced-order Itô Stochastic Differential Equation (ISDE) to sample from the learned distribution. However, its sampling dynamics are posed in the ambient space and the retained number of reduced coordinates is chosen by projection-reconstruction error. In practice, this often (i) requires more coordinates than the data’s intrinsic dimension to achieve stable sampling and (ii) lacks a smooth, basis-independent lifting back to the data domain; moreover, standard Diffusion Maps emphasize harmonic eigenfunctions and can miss non-harmonic latent structure. We address these limitations by decoupling geometry learning from sampling: a first Diffusion Maps pass identifies non-harmonic coordinates on which we formulate a full-order ISDE directly in the latent space, while Double Diffusion Maps captures multiscale geometric features and Geometric Harmonics (GH) learns a smooth lifting map to the ambient variables that is independent of the particular diffusion basis. This hybrid design preserves the system’s dynamical richness with a compact geometric representation and enables principled out-of-sample inference. The effectiveness and robustness of the proposed method are illustrated through two numerical studies: one based on data generated from two-dimensional Hermite polynomial functions and another based on high-fidelity simulations of a detonation wave in a reactive flow.

Double diffusion maps↗