Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Learning Algorithm”

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 1,567 records · Page 87
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

Learning Constrained Parametric Differentiable Predictive Control Policies With Guarantees

We present differentiable predictive control (DPC), a method for offline learning of constrained neural control policies for nonlinear dynamical systems with performance guarantees. We show that the sensitivities of the parametric optimal control problem can be used to obtain direct policy gradients. Specifically, we employ automatic differentiation (AD) to efficiently compute the sensitivities of the model predictive control (MPC) objective function and constraints penalties. To guarantee safety upon deployment, we derive probabilistic guarantees on closed-loop stability and constraint satisfaction based on indicator functions and Hoeffding’s inequality. We empirically demonstrate that the proposed method can learn neural control policies for various parametric optimal control tasks. In particular, we show that the proposed DPC method can stabilize systems with unstable dynamics, track time-varying references, and satisfy nonlinear state and input constraints. Our DPC method has practical time savings compared to alternative approaches for fast and memory-efficient controller design. Specifically, DPC does not depend on a supervisory controller as opposed to approximate MPC based on imitation learning. We demonstrate that, without losing performance, DPC is scalable with greatly reduced demands on memory and computation compared to implicit and explicit MPC while being more sample efficient than model-free reinforcement learning (RL) algorithms.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Graph learning for particle accelerator operations

Particle accelerators play a crucial role in scientific research, enabling the study of fundamental physics and materials science, as well as having important medical applications. This study proposes a novel graph learning approach to classify operational beamline configurations as good or bad. By considering the relationships among beamline elements, we transform data from components into a heterogeneous graph. We propose to learn from historical, unlabeled data via our self-supervised training strategy along with fine-tuning on a smaller, labeled dataset. Additionally, we extract a low-dimensional representation from each configuration that can be visualized in two dimensions. Leveraging our ability for classification, we map out regions of the low-dimensional latent space characterized by good and bad configurations, which in turn can provide valuable feedback to operators. This research demonstrates a paradigm shift in how complex, many-dimensional data from beamlines can be analyzed and leveraged for accelerator operations.

43 PARTICLE ACCELERATORS↗
NASA NTRS2009

Interface Generation and Compositional Verification in JavaPathfinder

We present a novel algorithm for interface generation of software components. Given a component, our algorithm uses learning techniques to compute a permissive interface representing legal usage of the component. Unlike our previous work, this algorithm does not require knowledge about the component s environment. Furthermore, in contrast to other related approaches, our algorithm computes permissive interfaces even in the presence of non-determinism in the component. Our algorithm is implemented in the JavaPathfinder model checking framework for UML statechart components. We have also added support for automated assume-guarantee style compositional verification in JavaPathfinder, using component interfaces. We report on the application of the presented approach to the generation of interfaces for flight software components.

Giannakopoulou, Dimitra↗
DOE OSTI2021

On Distributed Model-Free Reinforcement Learning Control with Stability Guarantee

Distributed learning can enable scalable and effective decision making in numerous complex cyber-physical systems such as smart transportation, robotics swarm, power systems, etc. However, the stability of the system is usually not guaranteed in most existing learning paradigms; and this limitation can hinder the wide deployment of machine learning in decision making of safety-critical systems. This paper presents a stability guaranteed distributed reinforcement learning (SGDRL) framework for interconnected linear subsystems, without knowing the subsystem models. While the learning process requires data from a peer-to-peer (p2p) communication architecture, the control implementation of each subsystem is only based on its local state. The stability of the interconnected subsystems will be ensured by a diagonally dominant eigenvalue condition, which will then be used in a model-free RL algorithm to learn the feedback control gains. The RL algorithm structure follows an off-policy iterative framework, with interleaved policy evaluation and policy update steps. We numerically validate our theoretical results by performing simulations on four interconnected sub-systems.

Mukherjee, Sayak↗
NASA NTRS2007

Salience Assignment for Multiple-Instance Regression

We present a Multiple-Instance Learning (MIL) algorithm for determining the salience of each item in each bag with respect to the bag's real-valued label. We use an alternating-projections constrained optimization approach to simultaneously learn a regression model and estimate all salience values. We evaluate this algorithm on a significant real-world problem, crop yield modeling, and demonstrate that it provides more extensive, intuitive, and stable salience models than Primary-Instance Regression, which selects a single relevant item from each bag.

regression↗
DOE OSTI2016

CLUSTERING-BASED FEATURE LEARNING ON VARIABLE STARS

The success of automatic classification of variable stars depends strongly on the lightcurve representation. Usually, lightcurves are represented as a vector of many descriptors designed by astronomers called features. These descriptors are expensive in terms of computing, require substantial research effort to develop, and do not guarantee a good classification. Today, lightcurve representation is not entirely automatic; algorithms must be designed and manually tuned up for every survey. The amounts of data that will be generated in the future mean astronomers must develop scalable and automated analysis pipelines. In this work we present a feature learning algorithm designed for variable objects. Our method works by extracting a large number of lightcurve subsequences from a given set, which are then clustered to find common local patterns in the time series. Representatives of these common patterns are then used to transform lightcurves of a labeled set into a new representation that can be used to train a classifier. The proposed algorithm learns the features from both labeled and unlabeled lightcurves, overcoming the bias using only labeled data. We test our method on data sets from the Massive Compact Halo Object survey and the Optical Gravitational Lensing Experiment; the results show that our classification performance is as good as and in some cases better than the performance achieved using traditional statistical features, while the computational cost is significantly lower. With these promising results, we believe that our method constitutes a significant step toward the automation of the lightcurve classification pipeline.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2022

Learning and Fast Adaptation for Grid Emergency Control via Deep Meta Reinforcement Learning

As power systems are undergoing a significant transformation with more uncertainties, less inertia and closer to operation limits, there is increasing risk of large outages. Thus, there is an imperative need to enhance grid emergency control to maintain system reliability and security. Towards this end, great progress has been made in developing deep reinforcement learning (DRL) based grid control solutions in recent years. However, existing DRL-based solutions have two main limitations: 1) they cannot handle well with a wide range of grid operation conditions, system parameters, and contingencies; 2) they generally lack the ability to fast adapt to new grid operation conditions, system parameters, and contingencies, limiting their applicability for real-world applications. Here, in this paper, we mitigate these limitations by developing a novel deep meta-reinforcement learning (DMRL) algorithm. The DMRL combines the meta strategy optimization together with DRL, and trains policies modulated by a latent space that can quickly adapt to new scenarios. We test the developed DMRL algorithm on the IEEE 300-bus system. We demonstrate fast adaptation of the meta-trained DRL polices with latent variables to new operating conditions and scenarios using the proposed method, which achieves superior performance compared to the state-of-the-art DRL and model predictive control (MPC) methods.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2018

Integrated Cyber/Physical Grid Resiliency Modeling

This project explored coupling modeling and analysis methods from multiple domains to address complex hybrid (cyber and physical) attacks on mission critical infrastructure. Robust methods to integrate these complex systems are necessary to enable large trade-space exploration including dynamic and evolving cyber threats and mitigations. Reinforcement learning employing deep neural networks, as in the AlphaGo Zero solution, was used to identify "best" (or approximately optimal) resilience strategies for operation of a cyber/physical grid model. A prototype platform was developed and the machine learning (ML) algorithm was made to play itself in a game of 'Hurt the Grid'. This proof of concept shows that machine learning optimization can help us understand and control complex, multi-dimensional grid space. A simple, yet high-fidelity model proves that the data have spatial correlation which is necessary for any optimization or control. Our prototype analysis showed that the reinforcement learning successfully improved adversary and defender knowledge to manipulate the grid. When expanded to more representative models, this exact type of machine learning will inform grid operations and defense - supporting mitigation development to defend the grid from complex cyber attacks! This same research can be expanded to similar complex domains.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2019

Quasi-Stochastic Approximation and Off-Policy Reinforcement Learning: Preprint

The Robbins-Monro stochastic approximation algorithm is a foundation of many algorithmic frameworks for reinforcement learning and often an efficient approach to solving (or approximating the solution to) complex optimal control problems. In many cases, however, practitioners are unable to apply these techniques because of an inherent high variance. This paper aims to provide a general foundation for 'quasi-stochastic approximation,' in which all of the processes under consideration are deterministic, much like quasi-Monte Carlo for variance reduction in simulation. The variance reduction can be substantial, subject to tuning of pertinent parameters in the algorithm. This paper introduces a new coupling argument to establish the optimal rate of convergence provided the gain is sufficiently large. These results are established for linear models and tested also in nonideal settings. A major application of these general results is a new class of reinforcement learning algorithms for deterministic state space models. In this setting, the main contribution is a class of algorithms for approximating the value function for a given policy, using a different policy designed to introduce exploration.

24 POWER TRANSMISSION AND DISTRIBUTION↗
NASA NTRS2009

Using Decision Trees to Detect and Isolate Simulated Leaks in the J-2X Rocket Engine

The goal of this work was to use data-driven methods to automatically detect and isolate faults in the J-2X rocket engine. It was decided to use decision trees, since they tend to be easier to interpret than other data-driven methods. The decision tree algorithm automatically "learns" a decision tree by performing a search through the space of possible decision trees to find one that fits the training data. The particular decision tree algorithm used is known as C4.5. Simulated J-2X data from a high-fidelity simulator developed at Pratt & Whitney Rocketdyne and known as the Detailed Real-Time Model (DRTM) was used to "train" and test the decision tree. Fifty-six DRTM simulations were performed for this purpose, with different leak sizes, different leak locations, and different times of leak onset. To make the simulations as realistic as possible, they included simulated sensor noise, and included a gradual degradation in both fuel and oxidizer turbine efficiency. A decision tree was trained using 11 of these simulations, and tested using the remaining 45 simulations. In the training phase, the C4.5 algorithm was provided with labeled examples of data from nominal operation and data including leaks in each leak location. From the data, it "learned" a decision tree that can classify unseen data as having no leak or having a leak in one of the five leak locations. In the test phase, the decision tree produced very low false alarm rates and low missed detection rates on the unseen data. It had very good fault isolation rates for three of the five simulated leak locations, but it tended to confuse the remaining two locations, perhaps because a large leak at one of these two locations can look very similar to a small leak at the other location.

Schwabacher, Mark A.↗
DOE OSTI2023

Intra-hour Solar Irradiance Forecast in Multiple Locations using Deep Transfer Learning

In recent years, solar power system installation imposes several challenges on the operations of local and regional power grids due to the inherent variability of ground-level solar irradiance. This work proposes a novel real-time solar forecast methodology for intra-hour solar irradiance based on deep transfer learning from ground-based sky imager for time horizons ranging from 5-15 min. There are three unique aspects of the proposed methodology: (1) a Deep Learning based algorithm development which is modeled as a classification approach rather than a traditional regression approach; (2) the use of the Transfer Learning technique to show generalization capability, robustness, and portability of baseline model in the newly deployed location where availability of enough data for training is typically scarce, and (3) redefinition of point-based irradiation forecast error estimation technique with a window-based one that is more intuitive and user-friendly. The system is developed using multiple years of irradiance and sky image recording in New Jersey and one-year data from Colorado, USA. The method is validated against ground telemetry from these two locations of diverse geographic and climatic conditions. Results show that the forecasting method proposed in this work is robust and highly accurate (8% MAPE error) for multiple locations deployment.

Deep Learning, Convolution Neural Networks, transf↗
DOE OSTI2022

MSU Disentanglement Analysis Software

This software is used to disentangle the forced-versus-unforced components of tropospheric temperature change over the satellite era (after 1979) using maps of surface temperature change as a predictor. In general, the software assembles training datasets (from pre-computed surface temperature trend maps and domain averaged tropospheric warming rates), trains statistical/machine learning (ML) algorithms, applies the trained statistical/ML model to climate model data and observations, and then saves the results. A leave-one-out approach is used in which the statistical/ML models are iteratively trained on (N- 1) climate model and then applied to the remaining climate model (and observations). Each model includes a large ensemble (i.e., >10) of model simulations. The software relies on scikit-learn ridge regression, PLS regression, and neural network algorithms.

Po-Chedley, StephenD↗
DOE OSTI2025

Optimizing Non-Terrestrial Hybrid RF/FSO Links With Reinforcement Learning: Navigating Through Clouds

In the pursuit of ubiquitous broadband connectivity, there has been a significant shift towards the vertical expansion of communication networks into space, particularly through the exploitation of low Earth orbit (LEO) satellite constellations, which are favored for their relatively low latency. However, this approach faces many challenges that need to be addressed, including atmospheric turbulence, high path loss, and dynamic cloud formations. High-altitude pseudo-satellites (HAPS) have emerged as promising relaying layers between LEO satellites and ground stations, enhancing coverage, latency, and direct terrestrial user connectivity. While radio frequency (RF) bands suffer from congestion and limited bandwidth, free space optical (FSO) communications offer higher data rates, but are susceptible to misalignment and weather-induced signal degradation. To address these challenges, a hybrid RF/FSO approach has been proposed to take advantage of both technologies by dynamic switching between RF and FSO based on propagation channel conditions. This paper introduces a reinforcement learning-based algorithm designed to optimize the trajectory of HAPS, maneuver around cloudy areas, and seamlessly switch between the RF and FSO communication modes to maximize the achievable capacity. The proposed approach aims to maximize system performance by intelligently adapting to environmental conditions and offering a promising solution for next-generation space communication networks.

actor-critic algorithm↗
DOE OSTI2018

Summit on HPC ML Algorithms and Human Systems Summary

The purpose of this document is to provide an initial overview of (new) information regarding the gaps in DOE's work on High Performance Computing(HPC), Machine Learning and Algorithms (IVILA) and Human Systems(HS) emerging from the Sep. 2018 Summit, and to briefly forecast the attendee community's next steps.

97 MATHEMATICS AND COMPUTING↗
NASA NTRS1994

Constructing Simplified Plans via Truth Criteria Approximation

This paper has presented an approach to dealing with the complexity of explanation-based learning plans in complex domains. This approach uses a simplified algorithm to construct plans, and employs later refinements to repair bugs in constructed plans. This algorithm has the theoretical properties of completeness and convergence upon soundness. This incremental reasoning planning and learning algorithm has been implemented using a partial-order constraint posting planner and empirically compared to a conventional exhaustive reasoning partial-order constraint-posting planner and learning algorithm. This comparison showed that 1) incremental reasoning significantly reduced learning costs compared to exhaustive reasoning, 2) Explanation-based Learning (EBL) reduced failures from incremental reasoning, 3) EBL with incremental reasoning required less search to solve problems than EBL with exhaustive reasoning.

Explanation-based learning EBL incremental reasoni↗
DOE OSTI2018

A multi-scale convolutional neural network for autonomous anomaly detection and classification in a laser powder bed fusion additive manufacturing process

In-situ detection of processing defects is a critical challenge for Laser Powder Bed Fusion Additive Manufacturing. Many of these defects are related to interactions between the recoater blade, which spreads the powder, and the powder bed. Here this work leverages Deep Learning, specifically a Convolutional Neural Network (CNN), for autonomous detection and classification of many of these spreading anomalies. Importantly, the input layer of the CNN is modified to enable the algorithm to learn both the appearance of the powder bed anomalies as well as key contextual information at multiple size scales. These modifications to the CNN architecture are shown to improve the flexibility and overall classification accuracy of the algorithm while mitigating many human biases. A case study is used to demonstrate the utility of the presented methodology and the overall performance is shown to be superior to that of methodologies previously reported by the authors.

36 MATERIALS SCIENCE↗
DOE OSTI2023

Plug-and-Play Methods for Integrating Physical and Learned Models in Computational Imaging: Theory, algorithms, and applications

Plug-and-play (PnP) priors constitute one of the most widely used frameworks for solving computational imaging problems through the integration of physical models and learned models. PnP leverages high-fidelity physical sensor models and powerful machine learning methods for prior modeling of data to provide state-of-the-art reconstruction algorithms. PnP algorithms alternate between minimizing a data fidelity term to promote data consistency and imposing a learned regularizer in the form of an image denoiser. Recent highly successful applications of PnP algorithms include biomicroscopy, computerized tomography (CT), magnetic resonance imaging (MRI), and joint ptychotomography. This article presents a unified and principled review of PnP by tracing its roots, describing its major variations, summarizing main results, and discussing applications in computational imaging. Additionally, we also point the way toward further developments by discussing recent results on equilibrium equations that formulate the problem associated with PnP algorithms.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Development of algorithms for augmenting and replacing conventional process control using reinforcement learning

Here, this work seeks to allow for the online operation and training of model-free reinforcement learning (RL) agents but limit the risk to system equipment and personnel. The parallel implementation of RL alongside more conventional process control (CPC) allows for the RL algorithm to learn from CPC. The past performance of both methods are assessed on a continuous basis allowing for a transition from CPC to RL and, if needed, transitioning back to CPC from RL. This allows for the RL algorithm to slowly and safely assume control of the process without significant degradation in control performance. It is shown that the RL can derive a near optimal policy even when coupled with a suboptimal CPC. It is also demonstrated that the coupled RL-CPC algorithm learns at a faster rate than traditional RL methods of exploration while the algorithm’s performance does not deteriorate below CPC, even when exposed to an unknown operating condition.

30 DIRECT ENERGY CONVERSION↗