Search NASA⌕ Search

SEARCH · Search NASA

Results for “Learning control”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

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

At least 1,297 records · Page 72
All sourcesNASA NTRSDOE OSTI
NASA NTRS2001

Using Model-Based Reasoning for Autonomous Instrument Operation - Lessons Learned From IMAGE/LENA

Model-based reasoning has been applied as an autonomous control strategy on the Low Energy Neutral Atom (LENA) instrument currently flying on board the Imager for Magnetosphere-to-Aurora Global Exploration (IMAGE) spacecraft. Explicit models of instrument subsystem responses have been constructed and are used to dynamically adapt the instrument to the spacecraft's environment. These functions are cast as part of a Virtual Principal Investigator (VPI) that autonomously monitors and controls the instrument. In the VPI's current implementation, LENA's command uplink volume has been decreased significantly from its previous volume; typically, no uplinks are required for operations. This work demonstrates that a model-based approach can be used to enhance science instrument effectiveness. The components of LENA are common in space science instrumentation, and lessons learned by modeling this system may be applied to other instruments. Future work involves the extension of these methods to cover more aspects of LENA operation and the generalization to other space science instrumentation.

Johnson, Michael A.↗
NASA NTRS2004

An Approach to V&V of Embedded Adaptive Systems

Rigorous Verification and Validation (V&V) techniques are essential for high assurance systems. Lately, the performance of some of these systems is enhanced by embedded adaptive components in order to cope with environmental changes. Although the ability of adapting is appealing, it actually poses a problem in terms of V&V. Since uncertainties induced by environmental changes have a significant impact on system behavior, the applicability of conventional V&V techniques is limited. In safety-critical applications such as flight control system, the mechanisms of change must be observed, diagnosed, accommodated and well understood prior to deployment. In this paper, we propose a non-conventional V&V approach suitable for online adaptive systems. We apply our approach to an intelligent flight control system that employs a particular type of Neural Networks (NN) as the adaptive learning paradigm. Presented methodology consists of a novelty detection technique and online stability monitoring tools. The novelty detection technique is based on Support Vector Data Description that detects novel (abnormal) data patterns. The Online Stability Monitoring tools based on Lyapunov's Stability Theory detect unstable learning behavior in neural networks. Cases studies based on a high fidelity simulator of NASA's Intelligent Flight Control System demonstrate a successful application of the presented V&V methodology. ,

Liu, Yan↗
DOE OSTI2022

A robust deep learning workflow to predict multiphase flow behavior during geological C O 2 sequestration injection and Post-Injection periods

Simulation of multiphase flow in porous media is essential to manage the geologic CO 2 sequestration (GCS) process, and physics-based simulation approaches usually take prohibitively high computational cost due to the nonlinearity of the coupled physics. This paper contributes to the development and evaluation of a deep learning workflow that accurately and efficiently predicts the temporal-spatial evolution of pressure and CO 2 plumes during injection and post-injection periods of GCS operations. Based on a Fourier Neural Operator, the deep learning workflow takes input variables or features including rock properties, well operational controls and time steps, and predicts the state variables of pressure and CO 2 saturation. To further improve the predictive fidelity, separate deep learning models are trained for CO 2 injection and post-injection periods due to the difference in primary driving force of fluid flow and transport during these two phases. We also explore different combinations of features to predict the state variables. We use a realistic example of CO 2 injection and storage in a 3D heterogeneous saline aquifer, and apply the deep learning workflow that is trained from physics-based simulation data and emulate the physics process. Through this numerical experiment, we demonstrate that using two separate deep learning models to distinguish post-injection from injection period generates the most accurate prediction of pressure, and a single deep learning model of the whole GCS process including the cumulative injection volume of CO 2 as a deep learning feature, leads to the most accurate prediction of CO 2 saturation. For the post-injection period, it is key to use cumulative CO 2 injection volume to inform the deep learning models about the total carbon storage when predicting either pressure or saturation. The deep learning workflow not only provides high predictive fidelity across temporal and spatial scales, but also offers a speedup of 250 times compared to full physics reservoir simulation, and thus will be a significant predictive tool for engineers to manage the long-term process of GCS.

58 GEOSCIENCES↗
DOE OSTI2019

Data-driven Feature Analysis in Control Design for Series-Compensated Transmission Systems

One challenge in power-system control designs is the gap between numerical model-based analysis and complex real-world power systems. With increased data and measurements being collected from power systems, data-driven analysis (e.g., machine learning) may provide an alternative approach to reveal hidden information through learning from the real system data, and provide insights for better control scheme design during the utility planning process. Data-driven feature analysis was performed to evaluate the relationships between series compensation, power generation, and path flows in a real transmission system, as well as temporal patterns. The main data-driven analysis methods, including statistical cross-correlation, multinomial logistical regression, and classification and regression trees, were integrated for feature selection and developing predictive models of series compensation. Analysis results demonstrated the effectiveness of the proposed methodology in feature analysis and the potential to help improve power-system control scheme design.

42 ENGINEERING↗
DOE OSTI2025

Grid Resiliency with a 100% Renewable Microgrid

San Diego Gas & Electric Company (SDG&E) installed America’s first and largest utility-scale microgrid in Borrego Springs in 2013. The first generation Borrego Springs Microgrid utilized diesel generators to form and stabilize the microgrid island, with support from grid-scale batteries and local solar photovoltaic (PV) generation. In this project, SDG&E in partnership with National Renewable Energy Laboratory (NREL) demonstrated through modeling, simulation and utility field testing that blackstart and islanding of the microgrid can be led with 100% renewable, inverter based resources (IBRs), to help reduce community reliance on conventional generation resources. Through equipment upgrades, grid-forming island leader capability was transitioned to a battery IBR instead of the Borrego Springs Microgrid diesel generators. A new microgrid controller was integrated to the microgrid and programmed to control and manage multiple energy storage systems. Synchrophasor and other power quality data verified autonomous, high-speed response of the IBRs through blackstart, islanding, and load step testing. Results of project field evaluations provide distribution systems operators (DSO) with increased confidence that renewable, IBR can replace traditional generators to blackstart and island microgrids and rapidly establish stable island frequency with rapid changes in peak power demand. Importantly, the project validated the integration feasibility of a distributed energy resource management system (DERMS) controller that manages multiple grid-forming and grid-following IBRs, establishing a standard design interface to reduce the complexity of integrating new DERs in the future and supporting replication by the industry. As a result of learnings in this project, SDG&E has implemented the microgrid controller strategy at multiple other microgrid sites, thereby validating the replicability of the solution. Hardware-in-the-loop (HIL) simulations including power and controller HIL hardware — along with electromagnetic transient (EMT) simulations of Borrego Springs Microgrid —informed adjustments to inverter parameters and were important to characterize the performance of the IBRs in relevant operating conditions before deployment. The EMT and HIL simulations of islanding the entire community are important contributions in providing confidence in IBR performance prior to future islanding of the community in the field. High-fidelity EMT and/or HIL simulation of IBRs can de-risk field operations, and its relevance and importance as a tool is increasing as distribution grids and microgrids become more complex and dynamic with an increasing proportion of renewable generation, distributed energy storage, and two-way power and energy flows.

24 POWER TRANSMISSION AND DISTRIBUTION↗
NASA NTRS2006

DCS-Neural-Network Program for Aircraft Control and Testing

A computer program implements a dynamic-cell-structure (DCS) artificial neural network that can perform such tasks as learning selected aerodynamic characteristics of an airplane from wind-tunnel test data and computing real-time stability and control derivatives of the airplane for use in feedback linearized control. A DCS neural network is one of several types of neural networks that can incorporate additional nodes in order to rapidly learn increasingly complex relationships between inputs and outputs. In the DCS neural network implemented by the present program, the insertion of nodes is based on accumulated error. A competitive Hebbian learning rule (a supervised-learning rule in which connection weights are adjusted to minimize differences between actual and desired outputs for training examples) is used. A Kohonen-style learning rule (derived from a relatively simple training algorithm, implements a Delaunay triangulation layout of neurons) is used to adjust node positions during training. Neighborhood topology determines which nodes are used to estimate new values. The network learns, starting with two nodes, and adds new nodes sequentially in locations chosen to maximize reductions in global error. At any given time during learning, the error becomes homogeneously distributed over all nodes.

Jorgensen, Charles C.↗
DOE OSTI2022

Towards multi-fidelity deep learning of wind turbine wakes

We report engineering wake models that accurately predict wake in a computationally efficient manner are very important for tasks such as layout optimization and control of wind farms. In this paper, we explore an application of deep learning (DL) to learn the wake model from hierarchies of physics-based approaches ranging from analytical models to an approximate form of the Reynolds-averaged Navier-Stokes equations. We first illustrate the application of principal component analysis to obtain a lower-dimensional representation that allows a computationally tractable training and deployment of DL models. Then, the DL model is trained to learn the mapping from input parameter space to the principal components, which are then used to reconstruct the three-dimensional flow field. Additionally, we investigate a composite framework consisting of two neural networks to learn the correlation between low- and high-fidelity data with Gauss and curl models treated as proxies for low- and high-fidelity models, respectively. The prediction from both DL models matches well with the high-fidelity data with a maximum relative percentage error for the kinetic energy flux of <1%. This work opens up possibilities for data-efficient construction of surrogate models for wake prediction that can be used to study the influence of wind speed and yaw angles on wind farm power production.

17 WIND ENERGY↗
NASA NTRS2006

Development and Flight Testing of a Neural Network Based Flight Control System on the NF-15B Aircraft

The Intelligent Flight Control System (IFCS) project at the NASA Dryden Flight Research Center, Edwards AFB, CA, has been investigating the use of neural network based adaptive control on a unique NF-15B test aircraft. The IFCS neural network is a software processor that stores measured aircraft response information to dynamically alter flight control gains. In 2006, the neural network was engaged and allowed to learn in real time to dynamically alter the aircraft handling qualities characteristics in the presence of actual aerodynamic failure conditions injected into the aircraft through the flight control system. The use of neural network and similar adaptive technologies in the design of highly fault and damage tolerant flight control systems shows promise in making future aircraft far more survivable than current technology allows. This paper will present the results of the IFCS flight test program conducted at the NASA Dryden Flight Research Center in 2006, with emphasis on challenges encountered and lessons learned.

Bomben, Craig R.↗
DOE OSTI2022

Recurrent neural networks for short-term and long-term prediction of geothermal reservoirs

Accurate prediction of geothermal reservoir responses to alternative energy production scenarios is critical for optimizing the development of the underlying resources. While the conventional physics-based models offer a comprehensive prediction tool, data-driven models provide an efficient alternative to build fit-for-purpose predictive models by extracting and using the statistical patterns in the collected data to make predictions. The recurrent neural network (RNN) is a data-driven model that is commonly applied to predict time series sequences. This paper presents a variant of RNN that also utilizes the efficiency of convolutional neural networks (CNN) for the prediction of energy production from geothermal reservoirs. Specifically, a CNN–RNN architecture is developed that takes historical well controls as input (features) and their corresponding production response data as output (labels) to learn an input-output mapping that can predict the future well production responses/performance for any given future well control inputs. The model is paired with a labeling scheme to handle real field disturbances that create data gaps. In addition to the model structure, we introduce a thorough workflow for applying the model, which includes data pre-processing, feature selection, as well as different training strategies for short-term and long-term prediction. Finally, the performance and accuracy of the model are evaluated by applying it to multiple datasets, including a field reservoir model.

15 GEOTHERMAL ENERGY↗
DOE OSTI2023

Creating a Simulation Platform for Research and Development of Advanced Control Methods

Advanced nuclear reactors are essential to meet the changing energy requirements throughout both the United States and the rest of world. In addition to other features, they are designed to enable deployment in remote locations and operate in a fully (or near-fully) autonomous manner, which will require a new control paradigm. To realize autonomously operating reactors, the U.S. Department of Energy’s Nuclear Energy Enabling Technologies Advanced Sensors and Instrumentation (NEET ASI) program conducts research and development into the enabling technologies and methods needed, including digital twins, machine learning, and risk modeling, in addition to various types of control methods. These technologies and methods are the key foundations needed to achieve fully autonomous systems. To develop and evaluate the technologies and methods necessary for achieving autonomous operations, it is critical to identify a software tool capable of integrating all the required elements. In surveying the available solutions, no software platforms were identified that could accomplish what was needed without introducing drawbacks. This challenge was the motivation for the current effort: to develop a software platform that can seamlessly integrate autonomouscontrol-enabling technologies and methods, allowing for accelerated research and development and transfer of ideas. The resulting platform, known as the Control and Optimization Modular Modeling Application for Nuclear Deployment (COMMAND), is Python-based, and leverages open-source tools to provide flexibility and facilitate building upon prior research. It is designed to enable advanced reactor developers to deploy and test advanced control technologies and methods coupled with their own models, solutions, and hardware. Given the substantial undertaking of developing such a platform, the current effort focused on laying down scalable, flexible software foundations and infrastructure, then demonstrating the platform via a use case. These foundations included developing generic modules, which contain the base variable and system blocks (the information and functional building blocks, respectively, that can be used to design a simulation) and the data handling and storage blocks needed to exchange information between the various blocks; as well as enablingtechnology-specific modules. This platform was evaluated via a use case, which was to simulate and control a process for the Microreactor Automated Control System (MACS) test bed. While MACS is not currently directly coupled to any specific microreactor physics, it was initially developed in concert with the Microreactor Applications Research Validation and Evaluation (MARVEL) microreactor, and so the MARVEL physics are used here. As part of this use case, several enabling-technology-specific blocks within COMMAND were integrated, including a proportional integral derivative (PID) control block, a Reactor Excursion and Leak Analysis Program (RELAP5-3D) block, and an anomaly detection block. The COMMAND software platform was successfully demonstrated to achieve the scalability and flexibility objectives of this effort and will be leveraged by the program’s research efforts to advance state of the art control methodologies towards autonomous operations of advanced reactors. As new use cases are created and implemented, it is anticipated that COMMAND will continue to grow and evolve to meet new requirements.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
DOE OSTI2023

Domain-aware Control-oriented Neural Models for Autonomous Underwater Vehicles

Conventional physics-based modeling is a time-consuming bottleneck in control design for complex nonlinear systems like autonomous underwater vehicles (AUVs). In contrast, purely data-driven models, require a large number of observations and lack operational guarantees for safety-critical systems. Data-driven models leveraging available partially characterized dynamics have potential to provide reliable systems models in a typical data-limited scenario for high value complex systems, thereby avoiding months of expensive expert modeling time. In this work we explore this middle-ground between expert-modeled and pure data-driven modeling. We present control-oriented parametric models with varying levels of domain-awareness that exploit known system structure and prior physics knowledge to create constrained deep neural dynamical system models. We employ universal differential equations to construct data-driven blackbox and graybox representations of the AUV dynamics. In addition, we explore a hybrid formulation that explicitly models the residual error related to imperfect graybox models. We compare the prediction performance of the learned models for different distributions of initial conditions and control inputs to assess their suitability for control.

Shaw Cortez, Wenceslao E.↗
DOE OSTI2025

Developing ML/AI Methods for High-Throughput Characterization of Multiple-Sensor Streams of Tokamak Dynamics for High-Speed Control (Final Report)

This project evaluated and developed new mathematical and algorithmic techniques capable of handling (in real-time) the growing amounts of data generated by modern fusion research. While existing numerical linear algebra (NLA) methods provide the backbone to classical data analysis and algorithms, these methods fundamentally do not port to distributed architectures nor do they allow low-latency data reduction for control. Motivated by the needs for modern fusion reactors, this project explored and implemented new numerical methods to characterize plasma dynamics, respond in real-time to discharge evolution, and to process massive-scale data accurately and rapidly more fully. This project links expertise in multiple-sensor diagnostics of tokamak plasma dynamics from Columbia University’s Plasma Physics Laboratory with expertise in massive-scale data reduction and extreme data control algorithms at Columbia University’s Data Science Institute. This interdisciplinary project (i) applied machine learning methods, (ii) implemented a properly-trained neural-network for very fast processing of high-speed plasma videography, and (ii) developed the applied mathematical methods, based on randomized-NLA (rNLA) routines, for data analysis, reduction, and real-time control. The Columbia University High Beta Tokamak-Extended Pulse (HBT-EP) facility provided data to test new algorithms and partnership with Columbia University's Data Sciences Institute evaluated the broader use of new algorithms for many challenging control applications.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2019

In Situ Quality Monitoring in AM Using Acoustic Emission: A Reinforcement Learning Approach

Additive manufacturing (AM) has attracted considerable attention in recent years. This technology overcomes the geometrical limits of workpieces produced with the traditional subtractive methods and so gives the opportunity to manufacture highly complex shapes. Unfortunately, the repeatability of the manufacturing process and the monitoring of quality are not reliable enough to be utilized in mass production. The quality monitoring of AM processes in commercial equipment has been largely based on temperature measurements of the process zone or high-resolution imaging of the layers. However, both techniques lack information about the physical phenomena taking place in the depth of the materials medium and this limits their reliability in real-life applications. To overcome those restrictions, we propose to combine acoustic emission and reinforcement learning. The former captures the information about the subsurface dynamics of the process. The latter is a branch of machine learning that allows interpreting the received data in terms of quality. The combination of both is an original method for in situ and real-time quality monitoring. Acoustic data were collected during a real process using a commercial AM machine. The process parameters were selected to achieve three levels of quality in terms of porosity concentration while manufacturing a stainless steel 316L cuboid shape. Using our method, we demonstrated that each level of quality produced unique acoustic signatures during the build that were recognized by the classifier. The classification accuracy reached in this work proves that the proposed method has high potential to be used as in situ and real-time monitoring of AM quality.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗
DOE OSTI2018

A digital twin for rapid qualification of 3D printed metallic components

The customized production of complex components by 3D printing has been hailed as a potentially transformative tool in manufacturing with important applications in health care, automotive and aerospace industries. However, after about a quarter of a century of research and development, only a handful of commercial alloys can be printed and the market value of all 3D printed products now amounts to a negligible portion of the manufacturing economy. This difficulty is attributable to a remarkable diversity in structure and properties of the printed components and susceptibility to defects. In addition, the current practice of qualifying components by prolonged trial and error with expensive printing equipment and feed stock material confine the printed products to a niche market where the high product cost and the delay in the qualification are not critical factors. In this report we explain how a digital twin or a digital replica of the printing machine will reduce the number of trial and error tests to obtain desired product attributes and reduce the time required for part qualification to make the printed components cost effective. It is shown that a comprehensive digital twin of 3D printing machine consisting of mechanistic, control and statistical models of 3D printing, machine learning and big data can reduce the volume of trial and error testing, reduce defects and shorten time between the design and production.

36 MATERIALS SCIENCE↗
DOE OSTI2021

Keras2c: A library for converting Keras neural networks to real-time compatible C

With the growth of machine learning models and neural networks in measurement and control systems comes the need to deploy these models in a way that is compatible with existing systems. Existing options for deploying neural networks either introduce very high latency, require expensive and time consuming work to integrate into existing code bases, or only support a very limited subset of model types. We have therefore developed a new method called Keras2c, which is a simple library for converting Keras/TensorFlow neural network models into real-time compatible C code. It supports a wide range of Keras layers and model types including multidimensional convolutions, recurrent layers, multi-input/output models, and shared layers. Keras2c re-implements the core components of Keras/TensorFlow required for predictive forward passes through neural networks in pure C, relying only on standard library functions considered safe for real-time use. The core functionality consists of ~ 1500 lines of code, making it lightweight and easy to integrate into existing codebases. Keras2c has been successfully tested in experiments and is currently in use on the plasma control system at the DIII-D National Fusion Facility at General Atomics in San Diego.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2025

Kelvin Probe Force Microscopy Imaging of Plasticity in Hydrogenated Perovskite Nickelate Multilevel Neuromorphic Devices

Ion drift in nanoscale electronically inhomogeneous semiconductors is among the most important mechanisms being studied for designing neuromorphic computing hardware. However, nondestructive imaging of the ion drift in operando devices directly responsible for multiresistance states and synaptic memory represents a formidable challenge. Here, we present Kelvin probe force microscopy imaging of hydrogen-doped perovskite nickelate device channels subject to high-speed electric field pulses to directly visualize proton distribution by monitoring surface potential changes spatially, which is also supported with finite element-based electric field distribution studies. First-principles calculations provide mechanistic insights into the origin of surface potential changes as a function of hydrogen donor doping that serves as the contrast mechanism. We demonstrate 128 (7-bit) nonvolatile conductance levels in such devices relevant to in-memory computing applications. The synaptic plasticity measurements are implemented in spiking neural networks and show promising results for classification (SciKit Learn’s Iris and Wine data sets) and control (OpenAI’s CartPole-v1 and BipedalWalker-v3) simulation tasks.

Kelvin probe force microscopy↗
DOE OSTI2023

Variational actor-critic algorithms,

We introduce a class of variational actor-critic algorithms based on a variational formulation over both the value function and the policy. The objective function of the variational formulation consists of two parts: one for maximizing the value function and the other for minimizing the Bellman residual. Besides the vanilla gradient descent with both the value function and the policy updates, we propose two variants, the clipping method and the flipping method, in order to speed up the convergence. We also prove that, when the prefactor of the Bellman residual is sufficiently large, the fixed point of the algorithm is close to the optimal policy.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

Enhancing SRF cavity stability and minimizing detuning with data-driven resonance control based on dynamic mode decomposition

Effective resonance control of superconducting radio frequency (SRF) cavities is critical for large machines like LCLS-II, as failure to achieve proper control can result in increased RF power consumption, higher cryogenic heat loads, and increased costs. To address this challenge, we have developed a machine learning (ML) model based on the dynamic mode decomposition method to represent the forced cavity dynamics. Using this model, we designed a model predictive controller (MPC) and demonstrated through simulation that the MPC can effectively stabilize the amplitude and phase of SRF cavities using only a frequency actuator, even in the presence of multiple mechanical modes. The lightweight and explicit ML model makes the controller suitable for direct implementation on field-programmable gate arrays, unlocking the full potential of SRF linacs like LCLS-II, enabling higher beam power and energy, and also serving as an advanced motion controller for various applications, such as photon beamlines and storage rings.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗