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 271 records · Page 15
All sourcesNASA NTRSDOE OSTI
DOE OSTI2022

Multi-Agent Deep Reinforcement Learning for Realistic Distribution System Voltage Control Using PV Inverters

Over the last few decades, the deployment of distributed solar photovoltaic (PV) systems has increased consistently. High PV penetration could cause adverse effects on the grid, such as voltage violations. This paper proposes a new distributed soft actor-critic based multi-agent deep reinforcement learning (SAC-MADRL) control solution to minimize the PV real power curtailment while keeping the grid voltage in an acceptable range. New reward functions have been designed to coordinate different agents during the learning process, yielding improved convergence. Comparison results with other control methods on a real feeder in western Colorado U.S. with 80% penetration of PVs demonstrate that the proposed method has better capability of effectively regulating voltage while minimizing the PV real power curtailment.

distribution system↗
DOE OSTI2021

Deep Reinforcement Learning for Autonomous Water Heater Control

Electric water heaters represent 14% of the electricity consumption in residential buildings. An average household in the United States (U.S.) spends about USD 400–600 (0.45 ¢/L–0.68 ¢/L) on water heating every year. In this context, water heaters are often considered as a valuable asset for Demand Response (DR) and building energy management system (BEMS) applications. To this end, this study proposes a model-free deep reinforcement learning (RL) approach that aims to minimize the electricity cost of a water heater under a time-of-use (TOU) electricity pricing policy by only using standard DR commands. In this approach, a set of RL agents, with different look ahead periods, were trained using the deep Q-networks (DQN) algorithm and their performance was tested on an unseen pair of price and hot water usage profiles. The testing results showed that the RL agents can help save electricity cost in the range of 19% to 35% compared to the baseline operation without causing any discomfort to end users. Additionally, the RL agents outperformed rule-based and model predictive control (MPC)-based controllers and achieved comparable performance to optimization-based control.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗
DOE OSTI2025

A Data Science and Machine Learning Platform Supporting Large Particle Accelerator Control and Diagnostics Applications Final Report: SBIR Initial Phase II DE-SC0022583

The Machine Learning Data Platform (MLDP) is a product providing full-stack support for data science, Machine Learning, and Artificial Intelligence (ML/AI) applications at particle accelerator and large experimental physics facilities. It supports ML/AI applications from front-end, high-speed acquisition of heterogeneous, time-series data, through data archiving and management, to back-end analysis. The MLDP embodies a “data-science ready” platform for data analysis and ML/AI applications in diagnosis, modelling, control, and optimization of these facilities. It provides data scientists and applications a consistent, datacentric interface to archive data standardizing implementation and deployment of ML/AI algorithms to different operations configurations within the same facility, or between facilities. Being an open-source, public-domain project, the MLDP is intended for broadest possible impact by increasing accessibility and minimizing the required expertise for installation and operation. The MLDP can also be deployed at user facilities for experimental data collection, archiving, and analysis. It is capable of acquisition and archiving of heterogeneous data from experimental equipment (e.g., images, arrays, structures, etc.) along with system hardware configurations (e.g., scalars, tables), control system process variables, and any metadata required for provenance. Thus, the MLDP can manage experimental data through its entire lifecycle, from acquisition and archiving, through analysis and investigation, to release and final publication.

43 PARTICLE ACCELERATORS↗
DOE OSTI2023

Deep Reinforcement Learning Based Smart Water Heater Control for Reducing Electricity Consumption and Carbon Emission

Water heating is the third largest electricity consumer in U.S. households, after space heating and cooling. Thus, water heaters represent a significant potential for reducing electricity consumption and associated CO2 emissions of residential buildings. To this end, this study proposes a model-free deep reinforcement learning (RL) approach that aims to minimize the electricity consumption and the CO2 emissions of a heat pump water heater without affecting user comfort. In this approach, a set of RL agents focusing on either electricity saving or emission reduction, with different look ahead periods, were trained using the deep Q-networks (DQN) algorithm and their performance was tested on different hot water usage and Marginal Operating Emissions Rate (MOER) profiles. The testing results showed that the RL agents that focus on electricity saving can save electricity in the range of 12–22% by operating the water heater with maximum heat pump efficiency and minimum electric element utilization. On the other hand, the RL agents that focus on emission reduction reduced emissions in the range of 18–37% by making use of the variable MOER values. These RL agents used the heat pump and/or an element when the MOER values are low due to the availability of renewable energy sources (e.g., solar and wind) and mostly avoided the periods of carbon-intensive periods. Overall, these results showed that the proposed RL approach can help minimize the electricity consumption and the CO2 emissions of a heat pump water heater without having any prior knowledge about the device.

Amasyali, Kadir↗
DOE OSTI2024

From Sim to Real: A Pipeline for Training and Deploying Traffic Smoothing Cruise Controllers

Designing and validating controllers for connected and automated vehicles to enhance traffic flow presents significant challenges, from the complexity of replicating real-world stop-and-go traffic dynamics in simulation, to the intricacies involved in transitioning from simulation to actual deployment. In this work, we present a full pipeline from data collection to controller deployment. Specifically, we collect 772 km of driving data from the I-24 in Tennessee, and use it to build a one-lane simulator, placing simulated vehicles behind real-world trajectories. Using policy-gradient methods with an asymmetric critic, we improve fuel efficiency by over 10% when simulating congested scenarios. Our comprehensive approach includes reinforcement learning for controller training, software verification, hardware validation and setup, and navigating various sim-to-real challenges. Furthermore, we analyze the controller's behavior and wave-smoothing properties, and deploy it on four Toyota Rav4’s in a real-world validation experiment on the I-24. Lastly, we release the driving dataset, the simulator and the trained controller, to enable future benchmarking and controller design.

42 ENGINEERING↗
DOE OSTI2021

Double Deep Q-Networks for Optimizing Electricity Cost of a Water Heater

Electric water heaters represent 14% of the electricity consumption in the residential buildings and the cost associated with domestic water heating account for a good portion of the household expenses in the United States. In this context, intelligent control of water heaters gained a lot of research attention. In recent years, a significant number of intelligent water heater controllers, with various methods and intended uses, have been proposed. However, existing studies are mostly model-based approaches that require an accurate modelling of the water heater. Towards addressing this research gap, this paper presents a model-free reinforcement learning-based controller for a day-ahead price market. The controller aims to minimize the cost of domestic water heating while maintaining the user comfort. The results showed that the developed controller can help save energy cost while maintaining the temperatures within the desired comfort band.

Amasyali, Kadir↗
DOE OSTI2023

Global-Local Policy Search and its Application in Grid-Interactive Building Control

As the buildings sector represents over 70% of the total U.S. electricity consumption, it offers a great amount of untapped demand-side resources to tackle many critical grid-side problems and improve the overall energy system's efficiency. To help make buildings grid-interactive, this paper proposes a global-local policy search method to train a reinforcement learning (RL) based controller which optimizes building operation during both normal hours and demand response (DR) events. Experiments on a simulated five-zone commercial building demonstrate that by adding a local fine-tuning stage to the evolution strategy policy training process, the control costs can be further reduced by 7.55% in unseen testing scenarios. Baseline comparison also indicates that the learned RL controller outperforms a pragmatic linear model predictive controller (MPC), while not requiring intensive online computation.

demand response↗
DOE OSTI2021

Machine Learning-Based Classification of Lignocellulosic Biomass from Pyrolysis-Molecular Beam Mass Spectrometry Data

High-throughput analysis of biomass is necessary to ensure consistent and uniform feedstocks for agricultural and bioenergy applications and is needed to inform genomics and systems biology models. Pyrolysis followed by mass spectrometry such as molecular beam mass spectrometry (py-MBMS) analyses are becoming increasingly popular for the rapid analysis of biomass cell wall composition and typically require the use of different data analysis tools depending on the need and application. Here, the authors report the py-MBMS analysis of several types of lignocellulosic biomass to gain an understanding of spectral patterns and variation with associated biomass composition and use machine learning approaches to classify, differentiate, and predict biomass types on the basis of py-MBMS spectra. Py-MBMS spectra were also corrected for instrumental variance using generalized linear modeling (GLM) based on the use of select ions relative abundances as spike-in controls. Machine learning classification algorithms e.g., random forest, k-nearest neighbor, decision tree, Gaussian Naïve Bayes, gradient boosting, and multilayer perceptron classifiers were used. The k-nearest neighbors (k-NN) classifier generally performed the best for classifications using raw spectral data, and the decision tree classifier performed the worst. After normalization of spectra to account for instrumental variance, all the classifiers had comparable and generally acceptable performance for predicting the biomass types, although the k-NN and decision tree classifiers were not as accurate for prediction of specific sample types. Gaussian Naïve Bayes (GNB) and extreme gradient boosting (XGB) classifiers performed better than the k-NN and the decision tree classifiers for the prediction of biomass mixtures. The data analysis workflow reported here could be applied and extended for comparison of biomass samples of varying types, species, phenotypes, and/or genotypes or subjected to different treatments, environments, etc. to further elucidate the sources of spectral variance, patterns, and to infer compositional information based on spectral analysis, particularly for analysis of data without a priori knowledge of the feedstock composition or identity.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2025

CRCNS US-France Research Proposal: Collaborative Research: Encoding reward expectation in Drosophilia

The fruit fly Drosophila melanogaster has been a valuable model for investigating the genetic and neural bases that underlie learning and memory. Early and most current studies use basic behavior conditioning protocols to study learning in controlled laboratory settings. More recently, the ability to transgenically manipulate many of the brain neurons in the fruit fly with exquisite specificity, and the recent knowledge of the synaptic ‘connectome’ of the fruit fly brain, makes these animals almost unique as a comprehensive model for studies of learning, memory and motivated behavior. In fact, the connectome has revealed many types of new connections that had until now been overlooked. Within this context, the thesis of this proposal is that studies of learning and memory will be greatly enhanced by using more sophisticated means for evaluating memory representations, such as have been developed in vertebrates, and combining those studies with information from the connectome guided by computational modelling. We propose to push beyond the boundaries of existing conditioning protocols for fruit flies to investigate more complex memory representations. In particular, we will investigate the function of reinforcement pathways in relation to the absence of expected reinforcement. More specifically, we propose a series of experiments designed to investigate the memory representations in fruit flies when an expected consequence of a Conditioned Stimulus (CS) fails to occur. Although studies have evaluated how this failure can establish extinction memory for the CS, our studies will go beyond studying extinction. Specifically, we predict that in Drosophila when a CS is associated with a failed expectation of an appetitive food reinforcement it will acquire aversive value, and vice versa for a failed expectation of an aversive reinforcer. We combine these studies with manipulations of reinforcement pathways in the CNS inspired from the connectome, iteratively knitted in with established computational models. Intellectual Merit: The concept of reinforcement expectation and incentive contrast have been influential in the development of studies of associative learning in mammals. These questions are particularly challenging to answer in vertebrates because they require exquisite cellular, temporal, and genetic specificity of experimental manipulations. The recent development of work with identified neurons and their connectomes makes the larval and adult fly brains ripe as models for pushing our understanding of neural bases for these higher- order conditioning phenomena. Broader Impacts: Public health: These analyses and the conceptual framework of prediction error processing underlying them have a profound impact on our understanding of reinforcement-related behavior in humans, including monetary rewards and the mnemonic consequences of traumatic experiences, and for pathologies of the dopamine reinforcement system. Educational: This project will provide interdisciplinary training for postdoctoral researchers, Ph.D. and undergraduate students. The PIs will act as co-supervisors or mentors of students working in the different labs via face-to-face and internet-based technologies. We will also work with ASU’s award-winning Ask- A-Biologist program. This is an online science program designed to enrich the learning experiences of students of all ages and to provide classroom material for use by K-12 teachers. We will develop an extension of a game developed under a prior NSF award, and the new game will include modules to teach K-12 students about how insects learn. We will also integrate into the AAB site a program developed by a collaborator (B Gerber) at the Leibniz Institut für Neurobiologie, Magdeburg, and now in use in schools in Germany, to teach K-12 students how to train animals using the fruit fly larval learning paradigm. Underrepresented groups: All PIs will work with their university offices of Academic Diversity and Equal Opportunity for reaching underrepresented students.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2021

Data-driven Optimal Control Strategy for Virtual Synchronous Generator via Deep Reinforcement Learning Approach

This paper aims at developing a data-driven optimal control strategy for virtual synchronous generator (VSG) in the scenario where no expert knowledge or requirement for system model is available. Firstly, the optimal and adaptive control problem for VSG is transformed into a reinforcement learning task. Specifically, the control variables, i.e., virtual inertia and damping factor, are defined as the actions. Meanwhile, the active power output, angular frequency and its derivative are considered as the observations. Moreover, the reward mechanism is designed based on three preset characteristic functions to quantify the control targets: (1) maintaining the deviation of angular frequency within special limits; (2) preserving well-damped oscillations for both the angular frequency and active power output; (3) obtaining slow frequency drop in the transient process. Next, to maximize the cumulative rewards, a decentralized deep policy gradient algorithm, which features model-free and faster convergence, is developed and employed to find the optimal control policy. With this effort, a data-driven adaptive VSG controller can be obtained. By using the proposed controller, the inverter-based distributed generator can adaptively adjust its control variables based on current observations to fulfill the expected targets in model-free fashion. Finally, simulation results validate the feasibility and effectiveness of the proposed approach.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2021

Data-Driven Multi-agent Deep Reinforcement Learning for Distribution System Decentralized Voltage Control with High Penetration of PVs

This paper proposes a novel model-free/data-driven centralized training and decentralized execution multi-agent deep reinforcement learning (MADRL) framework for distribution system voltage control with high penetration of PVs. The proposed MADRL can coordinate both the real and reactive power control of PVs with existing static var compensators and battery storage systems. Unlike the existing DRL-based voltage control methods, our proposed method does not rely on a system model during both the training and execution stages. This is achieved by developing a new interaction scheme between the surrogate modeling of the original system and the multi-agent soft actor critic (MASAC) MADRL algorithm. In particular, the sparse pseudo-Gaussian process with a few-shots of measurements is utilized to construct the surrogate model of the original environment, i.e., power flow model. This is a data-driven process and no model parameters are needed. Furthermore, the MASAC enabled MADRL allows to achieve better scalability by dividing the original system into different voltage control regions with the aid of real and reactive power sensitivities to voltage, where each region is treated as an agent. This also serves as the foundation for the centralized training and decentralized execution, thus significantly reducing the communication requirements as only local measurements are required for control. Comparative results with other alternatives on the IEEE 123-nodes and 342-nodes systems demonstrate the superiority of the proposed method.

14 SOLAR ENERGY↗
DOE OSTI2023

Implementing machine learning methods on QICK hardware for qubit readout & control

Quantum readout and control is a fundamental aspect of quantum computing that requires accurate measurement of qubit states. Errors emerge in all stages, from initialization to readout, and identifying errors in post-processing necessitates resource-intensive statistical analysis. In our work, we use a lightweight fully-connected neural network (NN) to classify states of a transmon system with no prior processing. Our NN accelerator yields higher fidelities (92%) than the classical matched filter method (84%). By exploiting the natural parallelism of NNs and their placement near the source of data on field-programmable gate arrays (FPGAs), we can achieve ultra-low latency on the Quantum Instrumentation Control Kit (QICK). Integrating machine learning methods on QICK opens several pathways for efficient real-time processing of quantum circuits.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2022

Dynamic Role-Based Access Control Policy for Smart Grid Applications: An Offline Deep Reinforcement Learning Approach

Role-based access control (RBAC) is adopted in the information and communication technology domain for authentication purposes. However, due to a very large number of entities within organizational access control (AC) systems, static RBAC management can be inefficient, costly, and can lead to cybersecurity threats. In this paper, a novel hybrid RBAC model is proposed, based on the principles of offline deep reinforcement learning (RL) and Bayesian belief networks. The considered framework utilizes a fully offline RL agent, which models the behavioral history of users as a Bayesian belief-based trust indicator. Thus, the initial static RBAC policy is improved in a dynamic manner through off-policy learning while guaranteeing compliance of the internal users with the security rules of the system. By deploying our implementation within the smart grid domain and specifically within a Distributed Energy Resources (DER) ecosystem, we provide an end-to-end proof of concept of our model. Finally, detailed analysis and evaluation regarding the offline training phase of the RL agent are provided, while the online deployment of the hybrid RL-based RBAC model into the DER ecosystem highlights its key operation features and salient benefits over traditional RBAC models.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗
DOE OSTI2020

Intelligent multi-zone residential HVAC control strategy based on deep reinforcement learning

Residential heating, ventilation, and air conditioning (HVAC) has been considered as an important demand response resource. However, the optimization of residential HVAC control is no trivial task due to the complexity of the thermal dynamic models of buildings and uncertainty associated with both occupant-driven heat loads and weather forecasts. In this paper, we apply a novel model-free deep reinforcement learning (RL) method, known as the deep deterministic policy gradient (DDPG), to generate an optimal control strategy for a multi-zone residential HVAC system with the goal of minimizing energy consumption cost while maintaining the users’ comfort. Here, the applied deep RL-based method learns through continuous interaction with a simulated building environment and without referring to any prior model knowledge. Simulation results show that compared with the state-of-art deep Q network (DQN), the DDPG-based HVAC control strategy can reduce the energy consumption cost by 15% and reduce the comfort violation by 79%; and when compared with a rule-based HVAC control strategy, the comfort violation can be reduced by 98%. In addition, experiments with different building models and retail price models demonstrate that the well-trained DDPG-based HVAC control strategy has high generalization and adaptability to unseen environments, which indicates its practicability for real-world implementation.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗
DOE OSTI2026

How Silica Surface Chemistry Modulates Interfacial Water: Insights from Machine Learning Molecular Dynamics

Controlling water structure and dynamics at silica interfaces are central to a wide range of technologies, including protective oxide layers for solar water splitting and nanoporous membranes. In this work, we develop a machine learning interatomic potential, trained via active learning, to achieve ab initio accuracy for water confined between hydroxylated silica surfaces over a range of silanol coverages and slit widths. We find that partially hydroxylated surfaces (50 and 75% OH) support stronger water−surface hydrogen bonding and more extended interfacial density profiles than fully hydroxylated (100% OH) surfaces, indicating that increasing OH coverage does not necessarily strengthen interfacial hydrogenbond networks. Translational diffusion decreases approximately linearly with slit width and OH coverage, whereas rotational dynamics respond nonlinearly. In particular, at the smallest slit width of 5 Å, 75% OH coverage produces an enhanced local tetrahedral ordered interfacial network that strongly suppresses reorientation, while 100% coverage yields a crowded, disordered interfacial layer that also hinders rotation. In contrast, the 50% OH coverage is sufficiently sparse that it does not markedly alter water structure or dynamics under confinement. These results show that coupled control of pore size and surface chemistry enables nonlinear tuning of interfacial water structure and transport, providing a design strategy for optimizing porous silica for either enhanced interfacial stability and controlled reactivity or rapid and selective transport.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2022

Deep Reinforcement Learning for Distribution System Restoration Using Distributed Energy Resources and Tie-Switches

Distributed energy resources (DERs), such as solar PVs and energy storage, can be used to restore distribution system critical loads after the extreme weather events to increase grid resilience. However, coordinating multiple DERs together with tie-switches for multi-step restoration process under renewable uncertainty is challenging. This paper proposes a deep reinforcement learning to control discrete actions of switching on/off tie switches and DERs for critical load restoration. The restoration problem is first cast into the Markov decision process suitable for DRL. Then, the original soft actor critic (SAC) method for continuous actions has been extended to handle discrete and continuous actions. Numerical comparison results with other stochastic optimization-based approaches on the modified IEEE 33-bus system show that the proposed method can achieve fast critical load restoration in the presence of substation power outage while maintaining system voltage limit throughout the restoration process.

active distribution systems↗
DOE OSTI2026

Feedback, physics, and forecasts: The emerging paradigm of machine learning-driven battery research

Machine learning (ML) is reshaping how we understand, predict, and optimize electrochemical systems. In batteries, ML accelerates discovery across chemistry, design, and operation by transforming massive experimental and simulated datasets into predictive, interpretable models. This review consolidates a decade of progress in ML-driven battery innovation, from early-cycle feature extraction to operando image analysis and physics-informed modeling. We categorize approaches by data domain and physical fidelity, emphasizing interpretable ML for diagnostics, reinforcement learning for control, and multi-objective optimization for lifetime extension strategies. Additionally, we demonstrate how integrated models accelerate discovery, reduce testing time, and guide sustainable design. Economic analyses furthermore illustrate how these advances can lower cost per cycle and improve circularity. Together, these developments chart a path toward self-optimizing, sustainable battery technologies.

artificial intelligence↗