Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “agent based”

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 1,441 records · Page 80
All sourcesNASA NTRSDOE OSTI
DOE OSTI2021

Restoring Critical Loads in Resilient Distribution Systems Using a Curriculum Learned Controller: Preprint

In this paper, we propose a curriculum learned reinforcement learning (RL) controller to facilitate distribution system critical load restoration (CLR), leveraging RL's fast online response and its outstanding optimal sequential control capability. Like many grid control problems, CLR is complicated due to the large control action space and renewable uncertainty in a heavily constrained non-linear environment with strong intertemporal dependency. The nature of the problem oftentimes causes the RL policy to converge to a poor-performing local optimum if learned directly. To overcome this, we design a two-stage curriculum in which the RL agent will learn generation control and load restoration decision under different scenarios progressively. Via curriculum learning, the trained RL controller is expected to achieve a better control performance, with critical loads restored as rapidly and reliably as possible. Using the IEEE 13-bus test system, we illustrate the performance of the RL controller trained by the proposed curriculum-based method.

61 RADIATION PROTECTION AND DOSIMETRY↗
DOE OSTI2020

Constrained Deep Reinforcement Learning for Energy Sustainable Multi-UAV Based Random Access IoT Networks With NOMA

In this paper, we apply the Non-Orthogonal Multiple Access (NOMA) technique to improve the massive channel access of a wireless IoT network where solar-powered Unmanned Aerial Vehicles (UAVs) relay data from IoT devices to remote servers. Specifically, IoT devices contend for accessing the shared wireless channel using an adaptive p-persistent slotted Aloha protocol; and the solar-powered UAVs adopt Successive Interference Cancellation (SIC) to decode multiple received data from IoT devices to improve access efficiency. To enable an energy-sustainable capacity-optimal network, we study the joint problem of dynamic multi-UAV altitude control and multi-cell wireless channel access management of IoT devices as a stochastic control problem with multiple energy constraints. We first formulate this problem as a Constrained Markov Decision Process (CMDP), and propose an online model-free Constrained Deep Reinforcement Learning (CDRL) algorithm based on Lagrangian primal-dual policy optimization to solve the CMDP. Extensive simulations demonstrate that our proposed algorithm learns a cooperative policy in which the altitude of UAVs and channel access probability of IoT devices are dynamically controlled to attain the maximal long-term network capacity while ensuring energy sustainability of UAVs, outperforming baseline schemes. The proposed CDRL agent can be trained on a small network, yet the learned policy can efficiently manage networks with a massive number of IoT devices and varying initial states, which can amortize the cost of training the CDRL agent.

42 ENGINEERING↗
DOE OSTI2025

eCounter: Inline Per-IP Network Monitoring at Millisecond Resolution via eBPF

Scientific data acquisition (SciDAQ) systems are shifting from archive-based workflows to streaming paradigms, where real-time, fine-grained network monitoring becomes essential. While P4-enabled devices offer per-packet in-band observability, they require specialized switches and routers. Host-side tools like Prometheus exporters lack sufficient temporal granularity. To bridge this gap, we present eCounter, a lightweight, hardware-agnostic, inline telemetry agent built on extended Berkeley Packet Filter (eBPF). eCounter captures per-interface ingress and egress traffic, categorized by IP address and protocol, at millisecond to sub-millisecond resolution. In a 100 Gbps environment, it continuously exports up to 3,257 time-series bins per second with only 4% CPU utilization at a 35¿KiB/s data rate. We evaluate eCounter across diverse NIC MTU settings, hook types, CPU architectures and operating systems, and observed negligible impact on concurrent high-throughput streaming applications. Complexity analysis confirms that it can be readily scaled to distributed SciDAQ deployments.

Mei, Xinxin [Computational Sciences and Technology↗
NASA NTRS1990

Cooperating systems: Layered MAS

Distributed intelligent systems can be distinguished by the models that they use. The model developed focuses on layered multiagent system conceived of as a bureaucracy in which a distributed data base serves as a central means of communication. The various generic bureaus of such a system is described and a basic vocabulary for such systems is presented. In presenting the bureaus and vocabularies, special attention is given to the sorts of reasonings that are appropriate. A bureaucratic model has a hierarchy of master system and work group that organizes E agents and B agents. The master system provides the administrative services and support facilities for the work groups.

Rochowiak, Daniel↗
NASA NTRS2000

Critical Heat Flux in Pool Boiling on Metal-Graphite Composite Surfaces

A study is conducted on high heat-flux pool boiling of pentane on micro-configured composite surfaces. The boiling surfaces are copper-graphite (Cu-Gr) and aluminum-graphite (Al-Gr) composites with a fiber volume concentration of 50%. The micro-graphite fibers embedded in the matrix contribute to a substantial enhancement in boiling heat-transfer performance. Correlation equations are obtained for both the isolated and coalesced bubble regimes, utilizing a mathematical model based on a metal-graphite, two-tier configuration with the aid of experimental data. A new model to predict the critical heat flux (CHF) on the composites is proposed to explain the fundamental aspects of the boiling phenomena. Three different factors affecting the CHF are considered in the model. Two of them are expected to become the main agents driving vapor volume detachment under microgravity conditions, using the metal-graphite composite surfaces as the heating surface and using liquids with an unusual Marangoni effect as the working fluid.

Zhang, Nengli↗
NASA NTRS2010

Characterization of Epoxy Functionalized Graphite Nanoparticles and the Physical Properties of Epoxy Matrix Nanocomposites

This work presents a novel approach to the functionalization of graphite nanoparticles. The technique provides a mechanism for covalent bonding between the filler and matrix, with minimal disruption to the sp2 hybridization of the pristine graphene sheet. Functionalization proceeded by covalently bonding an epoxy monomer to the surface of expanded graphite, via a coupling agent, such that the epoxy concentration was measured as approximately 4 wt.%. The impact of dispersing this material into an epoxy resin was evaluated with respect to the mechanical properties and electrical conductivity of the graphite-epoxy nanocomposite. At a loading as low as 0.5 wt.%, the electrical conductivity was increased by five orders of magnitude relative to the base resin. The material yield strength was increased by 30% and Young s modulus by 50%. These results were realized without compromise to the resin toughness.

Miller, Sandi G.↗
NASA NTRS2013

The Space Microbe Invasion: To Eat or Not to Eat

Objective: To investigate how different cleaning agents sanitize an assortment of vegetables and fruits for ~nsumption on board the International Space Station (ISS). -Description: This laboratory investigation will have students testing different cleaning agents on a variety of vegetables and fruits that can be grown on board the ISS. Students will determine which cleaning agent most effectively lowers the number of bacteria on a variety of vegetables and fruits. This lab will also lend itself to investigations dealing with pH and its' role in lowering bacterial counts. In addition, students will figure out the correct balance between plant architecture and effectiveness of sanitizing these surfaces to achieve lower bacteria counts. This will be determined based on swabbed bacteria samples later grown on a Petri dish.

Munoz, Angela↗
DOE OSTI2024

Overlooked cooling effects of albedo in terrestrial ecosystems

Radiative forcing (RF) resulting from changes in surface albedo is increasingly recognized as a significant driver of global climate change but has not been adequately estimated, including by Intergovernmental Panel on Climate Change (IPCC) assessment reports, compared with other warming agents. Here, we first present the physical foundation for modeling albedo-induced RF and the consequent global warming impact (GWI Δα ). We then highlight the shortcomings of available current databases and methodologies for calculating GWI Δα at multiple temporal scales. There is a clear lack of comprehensive in situ measurements of albedo due to sparse geographic coverage of ground-based stations, whereas estimates from satellites suffer from biases due to the limited frequency of image collection, and estimates from earth system models (ESMs) suffer from very coarse spatial resolution land cover maps and associated albedo values in pre-determined lookup tables. Field measurements of albedo show large differences by ecosystem type and large diurnal and seasonal changes. As indicated from our findings in southwest Michigan, GWI Δα is substantial, exceeding the RF Δα values of IPCC reports. Inclusion of GWI Δα to landowners and carbon credit markets for specific management practices are needed in future policies. We further identify four pressing research priorities: developing a comprehensive albedo database, pinpointing accurate reference sites within managed landscapes, refining algorithms for remote sensing of albedo by integrating geostationary and other orbital satellites, and integrating the GWI Δα component into future ESMs.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2022

Biochemical and structural insights into SARS-CoV-2 polyprotein processing by Mpro

SARS-CoV-2, a human coronavirus, is the causative agent of the COVID-19 pandemic. Its genome is translated into two large polyproteins subsequently cleaved by viral papain-like protease and main protease (Mpro). Polyprotein processing is essential yet incompletely understood. We studied Mpro-mediated processing of the nsp7-11 polyprotein, whose mature products include cofactors of the viral replicase, and identified the order of cleavages. Integrative modeling based on mass spectrometry (including hydrogen-deuterium exchange and cross-linking) and x-ray scattering yielded a nsp7-11 structural ensemble, demonstrating shared secondary structural elements with individual nsps. The pattern of cross-links and HDX footprint of the C145A Mpro and nsp7-11 complex demonstrate preferential binding of the enzyme active site to the polyprotein junction sites and additional transient contacts to help orient the enzyme on its substrate for cleavage. Last, proteolysis assays were used to characterize the effect of inhibitors/binders on Mpro processing/inhibition using the nsp7-11 polyprotein as substrate.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2026

TETA Autoresearch [SWR-26-089]

TETA Autoresearch is a template repository based on github.com/karpathy/autoresearch for AI-assisted research science in the TETA group in the Center for Integrated Mobility Sciences (CIMS) center at the National Laboratory of the Rockies. This software is a template for running autonomous research experiments that iteratively improve an ML model for a single optimization objective. Two execution modes share one harness: LLM mode - an agent (e.g. Claude Code) edits a scaffold train.py one change at a time, tagging each experiment, logging reasoning, and pushing results. Defined by program.md. Optimizer mode - an Optuna-backed driver (TPE / CMA-ES / Random) iterates over a domain-defined search space. Defined by optimizers/. RouteE (vehicle energy prediction) is the reference domain under domains/routee/. Adding a new domain is mechanical - see EXTENDING.md.

Reinicke, Nicholas [National Laboratory of the Roc↗
NASA NTRS1984

Applications of aerospace technology

The objective of the Research Triangle Institute Technology Transfer Team is to assist NASA in achieving widespread utilization of aerospace technology in terrestrial applications. Widespread utilization implies that the application of NASA technology is to benefit a significant sector of the economy and population of the Nation. This objective is best attained by stimulating the introduction of new or improved commercially available devices incorporating aerospace technology. A methodology is presented for the team's activities as an active transfer agent linking NASA Field Centers, industry associations, user groups, and the medical community. This methodology is designed to: (1) identify priority technology requirements in industry and medicine, (2) identify applicable NASA technology that represents an opportunity for a successful solution and commercial product, (3) obtain the early participation of industry in the transfer process, and (4) successfully develop a new product based on NASA technology.

Rouse, Doris J.↗
NASA NTRS1995

A decentralized software bus based on IP multicas ting

We describe decentralized reconfigurable implementation of a conference management system based on the low-level Internet Protocol (IP) multicasting protocol. IP multicasting allows low-cost, world-wide, two-way transmission of data between large numbers of conferencing participants through the Multicasting Backbone (MBone). Each conference is structured as a software bus -- a messaging system that provides a run-time interconnection model that acts as a separate agent (i.e., the bus) for routing, queuing, and delivering messages between distributed programs. Unlike the client-server interconnection model, the software bus model provides a level of indirection that enhances the flexibility and reconfigurability of a distributed system. Current software bus implementations like POLYLITH, however, rely on a centralized bus process and point-to-point protocols (i.e., TCP/IP) to route, queue, and deliver messages. We implement a software bus called the MULTIBUS that relies on a separate process only for routing and uses a reliable IP multicasting protocol for delivery of messages. The use of multicasting means that interconnections are independent of IP machine addresses. This approach allows reconfiguration of bus participants during system execution without notifying other participants of new IP addresses. The use of IP multicasting also permits an economy of scale in the number of participants. We describe the MULITIBUS protocol elements and show how our implementation performs better than centralized bus implementations.

Callahan, John R.↗
NASA NTRS2010

Fabrication and Modification of Nanoporous Silicon Particles

Silicon-based nanoporous particles as biodegradable drug carriers are advantageous in permeation, controlled release, and targeting. The use of biodegradable nanoporous silicon and silicon dioxide, with proper surface treatments, allows sustained drug release within the target site over a period of days, or even weeks, due to selective surface coating. A variety of surface treatment protocols are available for silicon-based particles to be stabilized, functionalized, or modified as required. Coated polyethylene glycol (PEG) chains showed the effective depression of both plasma protein adsorption and cell attachment to the modified surfaces, as well as the advantage of long circulating. Porous silicon particles are micromachined by lithography. Compared to the synthesis route of the nanomaterials, the advantages include: (1) the capability to make different shapes, not only spherical particles but also square, rectangular, or ellipse cross sections, etc.; (2) the capability for very precise dimension control; (3) the capacity for porosity and pore profile control; and (4) allowance of complex surface modification. The particle patterns as small as 60 nm can be fabricated using the state-of-the-art photolithography. The pores in silicon can be fabricated by exposing the silicon in an HF/ethanol solution and then subjecting the pores to an electrical current. The size and shape of the pores inside silicon can be adjusted by the doping of the silicon, electrical current application, the composition of the electrolyte solution, and etching time. The surface of the silicon particles can be modified by many means to provide targeted delivery and on-site permanence for extended release. Multiple active agents can be co-loaded into the particles. Because the surface modification of particles can be done on wafers before the mechanical release, asymmetrical surface modification is feasible. Starting from silicon wafers, a treatment, such as KOH dipping or reactive ion etching (RIE), may be applied to make the surface rough. This helps remove the nucleation layer. A protective layer is then deposited on the wafer. The protective layer, such as silicon nitride film or photoresist film, protects the wafer from electrochemical etching in an HF-based solution. A lithography technique is applied to pattern the particles onto the protective film. The undesired area of the protective film is removed, and the protective film on the back side of the wafer is also removed. Then the pattern is exposed to HF/surfactant solution, and a larger DC electrical current is applied to the wafers for a selected time. This step removes the nucleation layer. Then a DC current is applied to generate the nanopores. Next, a large electrical current is applied to generate a release layer. The particles are mechanically suspended in the solvent and collected by filtration or centrifuge.

Ferrari, Mauro↗
DOE OSTI2024

Synthesis and characterization of biobased copolyesters based on pentanediol: (2) Poly(pentylene adipate–co–terephthalate)

Traditionally, most flexible food packaging is made of linear low-density polyethylene (LLDPE) which cannot easily be recycled, nor will it degrade in a reasonable timescale. In this work, a biobased biodegradable polyester alternative was investigated as a possible replacement for LLDPE. High molecular weight poly (pentylene adipate-co-terephthalate) with a 40/60 adipic acid/terephthalic acid mole ratio was synthesized using direct esterification and polycondensation. Glycerol and hexane-1,2,5,6-tetrol were added as branching agents to better match the structure of the LLDPE which in turn might help the ability of these materials in film-blowing. Thermal, mechanical, and rheological properties of the copolyesters were thoroughly investigated. All copolyesters had a weight-average molecular weight of over 140,000 g/mol, which is necessary for proper rheology, and were thermally stable up to 350°C. Here, the addition of branching agents led to a slight decrease in crystallinity, d-spacing, melting temperature, enthalpy of melting, stress at break, and elongation at break. However, an increase in Young's modulus and complex viscosity at high frequency were observed compared to PPeAT60 without branching agent added. Although the improved crystallinity and mechanical properties of the copolyesters made them viable for film-blowing, the slow crystallization rate creates a major challenge.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2026

System-of-systems optimization of hydrogen infrastructure for heavy-duty freight corridors: The interstate 10 case study

Medium and heavy-duty freight transportation requires hydrogen energy infrastructure that is cost-effective, operationally reliable, spatially coherent, and resilient to demand variability along major corridors. This paper presents an integrated hydrogen corridor planning framework using Oak Ridge National Laboratory's OR-AGENT that couples freight-driven, route-resolved hydrogen demand modeling with optimized station siting, sizing, and station-level techno-economic analysis. The framework is demonstrated for the Interstate 10 freight corridor and the Houston-to-Los-Angeles region. Hydrogen demand is derived from high-resolution origin–destination freight data, duty-cycle characterization, and physics-based energy consumption modeling. Candidate refueling sites are selected from existing heavy-duty diesel fueling locations and optimized subject to onboard storage and station capacity constraints. Resulting station throughputs are evaluated using established techno-economic models for electrolytic hydrogen production and dispensing infrastructure. Results show that a regional, portfolio-level aggregation, average dispensed electrolytic hydrogen cost of $6.87–$7.26/kg is currently feasible, and is strongly influenced by demand density and utilization.

Sujan, Vivek [ORNL] (ORCID:0000000269882342)↗
DOE OSTI2022

Feedback Power Cost Optimization in Power Distribution Networks with Prosumers

Here, we propose a feedback control approach for solving optimal power flow (OPF) problems in power distribution networks (DNs) based on a projected gradient scheme, where the cost is given by the sum of functions of local power injections. This approach does not require detailed knowledge of the grid model and enables real-time tracking of the optimal generation configuration. Moreover, the communication burden is minimal: only a broadcast from a central entity, called the Network Supervisor (NS), to the agents in the network is needed. Hence, agents do not have to share local information that can be sensitive, e.g., their power injections or the value of the local costs. Our approach is then used to solve a particular OPF problem targeting DNs with high penetration of distributed energy resources (DERs) in which the energy cost is given by the remuneration that the DERs owners receive for the service they provide.

24 POWER TRANSMISSION AND DISTRIBUTION↗
DOE OSTI2024

Extreme Risk Mitigation in Reinforcement Learning using Extreme Value Theory

Risk-sensitive reinforcement learning (RL) has garnered significant attention in recent years due to the growing interest in deploying RL agents in real-world scenarios. A critical aspect of risk awareness involves modelling highly rare risk events (rewards) that could potentially lead to catastrophic outcomes. These infrequent occurrences present a formidable challenge for data-driven methods aiming to capture such risky events accurately. While risk-aware RL techniques do exist, they suffer from high variance estimation due to the inherent data scarcity. Our work proposes to enhance the resilience of RL agents when faced with very rare and risky events by focusing on refining the predictions of the extreme values predicted by the state-action value distribution. To achieve this, we formulate the extreme values of the state-action value function distribution as parameterized distributions, drawing inspiration from the principles of extreme value theory (EVT). We propose an extreme value theory based actor-critic approach, namely, Extreme Valued Actor-Critic (EVAC) which effectively addresses the issue of infrequent occurrence by leveraging EVT-based parameterization. Importantly, we theoretically demonstrate the advantages of employing these parameterized distributions in contrast to other risk-averse algorithms. Our evaluations show that the proposed method outperforms other risk averse RL algorithms on a diverse range of benchmark tasks, each encompassing distinct risk scenarios.

Wang, Yu↗