Search NASA⌕ Search

SEARCH · Search NASA

Results for “adaptive learning”

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 289 records · Page 16
All sourcesNASA NTRSDOE OSTI
NASA NTRS2021

A Learn-to-Fly Approach for Adaptively Tuning Flight Control Systems

A method is presented for adaptively tuning feedback control gains in a flight control system to achieve desired closed-loop performance. The method combines efficient aircraft parameter estimation, for identifying closed-loop dynamics models, with online nonlinear optimization, for sequentially perturbing and updating control gains to improve performance. Access to flight measurements and the control gains is required, but no prior information about the aircraft dynamics or the flight control architecture is needed. After the procedure, the optimized control gains (with uncertainties), the open-loop dynamics model, and the closed-loop dynamics model are available. The method is demonstrated for tuning a longitudinal stability augmentation system using a realistic nonlinear flight dynamics simulation of a subscale airplane. Convergence was attained using five maneuvers that spanned approximately one minute of flight test time. Although shown for a relatively simple case, the method is general and can be applied to other aircraft, axes of motion, performance metrics, and control system designs.

Adaptive tuning↗
DOE OSTI2016

shiftNMFk 1.1: Robust Nonnegative matrix factorization with kmeans clustering and signal shift, for allocation of unknown physical sources, toy version for open sourcing with publications

This code is a toy (short) version of CODE-2016-83. From a general perspective, the code represents an unsupervised adaptive machine learning algorithm that allows efficient and high performance de-mixing and feature extraction of a multitude of non-negative signals mixed and recorded by a network of uncorrelated sensor arrays. The code identifies the number of the mixed original signals and their locations. Further, the code also allows deciphering of signals that have been delayed in regards to the mixing process in each sensor. This code is high customizable and it can be efficiently used for a fast macro-analyses of data. The code is applicable to a plethora of distinct problems: chemical decomposition, pressure transient decomposition, unknown sources/signal allocation, EM signal decomposition. An additional procedure for allocation of the unknown sources is incorporated in the code.

Alexandrov, Boian S.↗
DOE OSTI2017

A Targeted Attack For Enhancing Resiliency of Intelligent Intrusion Detection Modules in Energy Cyber Physical Systems

Abstract— Secure high-speed communication is required to ensure proper operation of complex power grid systems and prevent malicious tampering activities. In this paper, artificial neural networks with temporal dependency are introduced for false data identification and mitigation for broadcasted IEC 61850 SMV messages. The fast responses of such intelligent modules in intrusion detection make them suitable for time- critical applications, such as protection. However, care must be taken in selecting the appropriate intelligence model and decision criteria. As such, this paper presents a customizable malware script to sniff and manipulate SMV messages and demonstrates the ability of the malware to trigger false positives in the neural network’s response. The malware developed is intended to be as a vaccine to harden the intrusion detection system against data manipulation attacks by enhancing the neural network’s ability to learn and adapt to these attacks.

25 ENERGY STORAGE↗
DOE OSTI2021

Protection of Radial Circuits with High Penetration Distributed Photovoltaics

The paper addresses the shortcomings of distribution protection systems as distribution level photovoltaic installations as well as other inverter based distributed resources are added to the system. Several protection schemes have been evaluated, including distance-based, estimation-based and learning-based. The paper summarizes the results and provides recommendations.

distance protection, radial distribution feeders, ↗
DOE OSTI2019

Robot Adaptation to Unstructured Terrains by Joint Representation and Apprenticeship Learning

When a mobile robot is deployed in a field environment, e.g., during a disaster response application, the capability of adapting its navigational behaviors to unstructured terrains is essential for effective and safe robot navigation. In this paper, we introduce a novel joint terrain representation and apprenticeship learning approach to implement robot adaptation to unstructured terrains. Different from conventional learning-based adaptation techniques, our approach provides a unified problem formulation that integrates representation and apprenticeship learning under a unified regularized optimization framework, instead of treating them as separate and independent procedures. Our approach also has the capability to automatically identify discriminative feature modalities, which can improve the robustness of robot adaptation. In addition, we implement a new optimization algorithm to solve the formulated problem, which provides a theoretical guarantee to converge to the global optimal solution. In the experiments, we extensively evaluate the proposed approach in real-world scenarios, in which a mobile robot navigates on familiar and unfamiliar unstructured terrains. Experimental results have shown that the proposed approach is able to transfer human expertise to robots with small errors, achieve superior performance compared with previous and baseline methods, and provide intuitive insights on the importance of terrain feature modalities.

42 ENGINEERING↗
DOE OSTI2023

Off-policy deep reinforcement learning with automatic entropy adjustment for adaptive online grid emergency control

Electric overloading conditions and contingencies put modern power systems at risk of voltage collapse and blackouts. Load shedding is crucial to maintain voltage stability for grid emergency control. However, the rule- or model-based schemes rely on accurate dynamic system models and face considerable challenges in adapting to various operating conditions and uncertain event occurrences. Here, to address these issues, this paper proposes a novel deep reinforcement learning (DRL)-based voltage stability control algorithm with automatic entropy adjustment (AEA) for grid emergency control. Various dynamic network components for complex system operations are modeled to construct the DRL environment. An off-policy soft actor-critic architecture is developed to maximize the expected reward and policy entropy simultaneously. The AEA mechanism is proposed to facilitate the policy maximum entropy procedure, and the proposed method can automatically provide effective discrete and continuous actions against various fault scenarios. Our approach accomplishes high sampling efficiency, scalability, and auto-adaptivity of the control policies under high uncertainties. Comparative studies with the existing DRL-based control methods in IEEE benchmarks indicate salient performance improvement of the proposed method for dynamic system emergency control.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2022

Learning high-dimensional parametric maps via reduced basis adaptive residual networks

We propose a scalable framework for the learning of high-dimensional parametric maps via adaptively constructed residual network (ResNet) maps between reduced bases of the inputs and outputs. When just few training data are available, it is beneficial to have a compact parametrization in order to ameliorate the ill-posedness of the neural network training problem. By linearly restricting high-dimensional maps to informed reduced bases of the inputs, one can compress high-dimensional maps in a constructive way that can be used to detect appropriate basis ranks, equipped with rigorous error estimates. A scalable neural network learning framework is thus to learn the nonlinear compressed reduced basis mapping. Unlike the reduced basis construction, however, neural network constructions are not guaranteed to reduce errors by adding representation power, making it difficult to achieve good practical performance. Inspired by recent approximation theory that connects ResNets to sequential minimizing flows, we present an adaptive ResNet construction algorithm. This algorithm allows for depth-wise enrichment of the neural network approximation, in a manner that can achieve good practical performance by first training a shallow network and then adapting. We prove universal approximation of the associated neural network class for $L^2_v$ functions on compact sets. Our overall framework allows for constructive means to detect appropriate breadth and depth, and related compact parametrizations of neural networks, significantly reducing the need for architectural hyperparameter tuning. Numerical experiments for parametric PDE problems and a 3D CFD wing design optimization parametric map demonstrate that the proposed methodology can achieve remarkably high accuracy for limited training data, and outperformed other neural network strategies we compared against.

42 ENGINEERING↗
DOE OSTI2019

Clutter Identification based on sparse recovery with dynamically changing dictionary sizes for cognitive radar

Existing radar algorithms assume stationary statistical characteristics for environment/clutter. In practical scenarios, the statistical characteristics of the clutter can dynamically change depending on where the radar is operating. Non-stationarity in the statistical characteristics of the clutter may negatively affect the radar performance. Cognitive radar that can sense the changes in the clutter statistics, learn the new statistical characteristics, and adapt to these changes has been proposed to overcome these shortcomings. We have recently developed techniques for detection of statistical changes and learning the new clutter distribution for cognitive radar. In this work, we will extend the learning component. More specifically, in our previous work, we have developed a sparse recovery based clutter distribution identification to learn the distribution of the new clutter characteristics after the detected change in the statistics of the clutter. In our method, we have built a dictionary of clutter distributions and used this dictionary in orthogonal matching pursuit to perform sparse recovery of the clutter distribution assuming that the dictionary includes the new distribution. In this work, we propose a hypothesis testing based approach to detect whether the new distribution of the clutter is included in the dictionary or not, and suggest a method to dynamically update the dictionary. We envision that the successful outcomes of this work will be of high relevance to the adaptive learning and cognitive augmentation of the radar systems that are used in remotely piloted vehicles for surveillance and reconnaissance operations.

Zhu, Yuansheng↗
DOE OSTI2015

Active Learning in the Era of Big Data

Active learning methods automatically adapt data collection by selecting the most informative samples in order to accelerate machine learning. Because of this, real-world testing and comparing active learning algorithms requires collecting new datasets (adaptively), rather than simply applying algorithms to benchmark datasets, as is the norm in (passive) machine learning research. To facilitate the development, testing and deployment of active learning for real applications, we have built an open-source software system for large-scale active learning research and experimentation. The system, called NEXT, provides a unique platform for realworld, reproducible active learning research. This paper details the challenges of building the system and demonstrates its capabilities with several experiments. The results show how experimentation can help expose strengths and weaknesses of active learning algorithms, in sometimes unexpected and enlightening ways.

Jamieson, Kevin↗
DOE OSTI2023

Learning to learn by using nonequilibrium training protocols for adaptable materials

Evolution in time-varying environments naturally leads to adaptable biological systems that can easily switch functionalities. Advances in the synthesis of environmentally responsive materials therefore open up the possibility of creating a wide range of synthetic materials which can also be trained for adaptability. We consider high-dimensional inverse problems for materials where any particular functionality can be realized by numerous equivalent choices of design parameters. By periodically switching targets in a given design algorithm, we can teach a material to perform incompatible functionalities with minimal changes in design parameters. We exhibit this learning strategy for adaptability in two simulated settings: elastic networks that are designed to switch deformation modes with minimal bond changes and heteropolymers whose folding pathway selections are controlled by a minimal set of monomer affinities. The resulting designs can reveal physical principles, such as nucleation-controlled folding, that enable such adaptability.

Science & Technology - Other Topics↗
DOE OSTI2024

Adaptive Quantum Generative Training using an Unbounded Loss Function

We propose a generative quantum learning algorithm using the Adaptive Derivative-Assembled Problem Tailored ansatz (ADAPT) framework in which the loss function to be minimized is the maximal quantum Rényi divergence of order two, an unbounded function that mitigates barren plateaus which inhibit training variational circuits. We benchmark this method against other state-of-the-art adaptive algorithms by learning random two-local thermal states. We perform numerical experiments of up to 12 qubits comparing our method learning algorithms that use linear objective functions and show that Rényi-ADAPT is capable of constructing shallow quantum circuits competitive with existing methods, while the gradients remain favorable resulting from the maximal Rényi divergence loss function.

quantum algorithms, quantum machine learning, quan↗
DOE OSTI2025

Estimating the Impacts of Increasing Temperatures and the Efficacy of Climate Adaptation Strategies in Urban Microclimates with Deep Learning

As urbanization and climate change progress, understanding and addressing urban heat becomes a priority for climate adaptation efforts. High temperatures concentrated in the urban core can drive increased risk of heat-related death and illness as well as increased energy demand for cooling. However, modeling the urban microclimate is an ongoing field of research typically burdened by an imprecise description of the built environment, incomplete observational records, significant computational cost, and a lack of high-resolution estimates of the impacts of increasing temperatures. Here, we present computationally efficient machine learning methods that can improve the accuracy of urban temperature estimates when compared to historical reanalysis data. These models are applied to a neighborhood in Los Angeles, and we compare the energy benefits of heat mitigation strategies to the impacts of climate change. We find that cooling demand is likely to increase substantially through midcentury, but engineered high-albedo surfaces could lessen this increase by more than 50 %. The corresponding increase in winter gas heating offsets the summer cooling benefit in the current climate, but total annual energy use from combined heating and cooling with electric heat pumps benefits from the engineered heat mitigation strategies under both current and future climates.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2017

Adaptive Neuron Apoptosis for Accelerating Deep Learning on Large Scale Systems

Machine Learning and Data Mining (MLDM) algorithms are becoming ubiquitous in {\em model learning} from the large volume of data generated using simulations, experiments and handheld devices. Deep Learning algorithms -- a class of MLDM algorithms -- are applied for automatic feature extraction, and learning non-linear models for unsupervised and supervised algorithms. Naturally, several libraries which support large scale Deep Learning -- such as TensorFlow and Caffe -- have become popular. In this paper, we present novel techniques to accelerate the convergence of Deep Learning algorithms by conducting low overhead removal of redundant neurons -- {\em apoptosis} of neurons -- which do not contribute to model learning, during the training phase itself. We provide in-depth theoretical underpinnings of our heuristics (bounding accuracy loss and handling apoptosis of several neuron types), and present the methods to conduct adaptive neuron apoptosis. We implement our proposed heuristics with the recently introduced TensorFlow and using its recently proposed extension with MPI. Our performance evaluation on two difference clusters -- one connected with Intel Haswell multi-core systems, and other with nVIDIA GPUs -- using InfiniBand, indicates the efficacy of the proposed heuristics and implementations. Specifically, we are able to improve the training time for several datasets by 2-3x, while reducing the number of parameters by 30x (4-5x on average) on datasets such as ImageNet classification. For the Higgs Boson dataset, our implementation improves the accuracy (measured by Area Under Curve (AUC)) for classification from 0.88/1 to 0.94/1, while reducing the number of parameters by 3x in comparison to existing literature, while achieving a 2.44x speedup in comparison to the default (no apoptosis) algorithm.

Siegel, Charles M.↗
DOE OSTI2026

Sim-to-real supervised domain adaptation for radioisotope identification

Machine learning has the potential to improve the speed and reliability of radioisotope identification using gamma spectroscopy. However, meticulously labeling an experimental dataset for training is often prohibitively expensive, while training models purely on synthetic data is risky due to the domain gap between simulated and experimental measurements. In this research, we demonstrate that supervised domain adaptation can substantially improve the performance of radioisotope identification models by transferring knowledge between synthetic and experimental data domains. We consider two domain adaptation scenarios: (1) a simulation-to-simulation adaptation, where we perform multi-label proportion estimation using simulated high-purity germanium detectors, and (2) a simulation-to-experimental adaptation, where we perform multi-class, single-label classification using measured spectra from handheld lanthanum bromide (LaBr) and sodium iodide (NaI) detectors. We begin by pretraining a spectral classifier on synthetic data using a custom transformer-based neural network. After subsequent fine-tuning on just 64 labeled experimental spectra, we achieve a test accuracy of 96% in the sim-to-real scenario with a LaBr detector, far surpassing a synthetic-only baseline model (75%) and a model trained from scratch (80%) on the same 64 spectra. Furthermore, we demonstrate that domain-adapted models learn more human-interpretable features than experiment-only baseline models. Overall, our results highlight the potential for supervised domain adaptation techniques to bridge the sim-to-real gap in radioisotope identification, enabling the development of accurate and explainable classifiers even in real-world scenarios where access to experimental data is limited.

Lalor, Peter W.↗
DOE OSTI2025

Balancing Trade-offs: Adaptive Differential Privacy in Interpretable Machine Learning Models

In the advancing field of machine learning, balancing accuracy, interpretability, and privacy represents a significant challenge. The problem is exacerbated by the widespread deployment of pre-trained models locally in diverse applications, which could lead to various amounts of privacy leakage. Conventional Differential Privacy strategies, in which uniform noises are applied to model gradients, guarantee data privacy at the expense of accuracy and interpretability. This paper introduces a Feature-Sensitive Adaptive Differential Privacy (FADP) framework with a unique noise-adding strategy. Noises are adaptively added based on feature importance clustering, where important features are considered for interpretability. By employing a unique masking technique, FADP selectively preserves crucial features with minimal noise interference, maintaining accuracy while enhancing interpretability. The FADP framework addresses the limitations of traditional DP methods by preserving critical channels and improving interpretability — a vital requirement in machine learning applications that demand transparency in model decisions. Through comprehensive testing, FADP is shown to balance the trade-offs among accuracy, privacy, and interpretability, marking a substantial advancement in the field of privacy-preserving machine learning.

Farhad Riya, Farhin [University of Tennessee, Knox↗