Search NASA⌕ Search

SEARCH · Search NASA

Results for “operation window”

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 307 records · Page 17
All sourcesNASA NTRSDOE OSTI
DOE OSTI2022

Scattered-light uniformity imager for diagnosing laser absorption asymmetries on OMEGA

Light scattered from a target is the most-direct measurement for diagnosing laser absorption in a direct-drive implosion. Observations from OMEGA implosions show much larger scattered-light asymmetries than predictions. A new instrument has been developed to absolutely measure the scattered-light intensity and nonuniformity for the purpose of diagnosing the asymmetry. The scattered-light uniformity imager (SLUI) diagnostic records the variation in scattered-light intensity over a transmission diffuser using a charge-coupled device (CCD)/lens assembly. At the standard operating position, an 11.3° (f/2.5) cone of light is collected. A stray light baffle, debris shield, and antireflection absorbing filter are also incorporated into the diagnostic payload inserted into the target chamber. The imaging parts of the diagnostic (light baffle, vacuum window, filters, lens, and CCD camera) are located outside the target chamber. Five SLUIs have been built and deployed in OMEGA’s ten-inch manipulator diagnostic ports, covering almost 5% of the emission surface, enabling an absolute scattered-light measurement should be within a few percent of the global average. Each SLUI system is calibrated offline, providing absolute scattered-light intensity measurements. Based on the measured point spread function, each diffuser plate image contains over 20 000 independent scattered-light absolute-intensity measurements of the variation over the collection cone. SLUI provides a platform to study scattered light and absorption asymmetries, and their possible sources.

47 OTHER INSTRUMENTATION↗
DOE OSTI2025

Network analysis of memristive device circuits: dynamics, stability and correlations

Abstract Networks with memristive devices are a potential basis for the next generation of computing devices. They are also an important model system for basic science, from modeling nanoscale conductivity to providing insight into the information-processing of neurons. The resistance in a memristive device depends on the history of the applied bias and thus displays a type of memory. The interplay of this memory with the dynamic properties of the network can give rise to new behavior, offering many fascinating theoretical challenges. But methods to analyze general memristive circuits are not well described in the literature. In this paper we develop a general circuit analysis for networks that combine memristive devices alongside resistors, capacitors and inductors and under various types of control. We derive equations of motion for the memory parameters of these circuits and describe the conditions for which a network should display properties characteristic of a resonator system. For the case of a purely memresistive network, we derive Lyapunov functions, which can be used to study the stability of the network dynamics. Surprisingly, analysis of the Lyapunov functions show that these circuits do not always have a stable equilibrium in the case of nonlinear resistance and window functions. The Lyapunov function allows us to study circuit invariances, wherein different circuits give rise to similar equations of motion, which manifest through a gauge freedom and node permutations. Finally, we identify the relation between the graph Laplacian and the operators governing the dynamics of memristor networks operators, and we use these tools to study the correlations between distant memristive devices through the effective resistance.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2022

Defining the Prototype Convenience Store Building Characteristics

The US Department of Energy supports the development of commercial building energy codes and standards. To support commercial building energy research activities and the development of commercial building energy codes and standards, continuous efforts have been made to convert 30 prototype building models (20 different building types), which cover 80% of US commercial building floor space, to OpenStudio prototype buildings. Additionally, the suite of prototype building models was expanded to include the addition of new building prototype models (e.g., courthouse, college building). This report documents the building and system characteristics of the prototype convenience store building model. Multiple sources, including databases, documented projects, and personal communications, were used to define the prototype convenience store building characteristics, and a one-story, 3,000 ft 2 building was considered as the prototype convenience store to represent an average-sized convenience store in the United States. The Oak Ridge National Laboratory team defined the operational hours of the convenience store as 17 hours per day. The types of exterior walls, roof, and floor were defined as brick wall, metal surfacing roof, and slab-on-grade floor, respectively. Double glazing was selected, and the window-towall ratio was set as 15% on the front side. For the system part, packaged system and centralized water heater were selected for the space heating and cooling system and water heater system, respectively. In terms of the refrigerator system, the number of refrigerators and freezers were determined as eight closed cases and two walk-in units.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗
DOE OSTI2025

Exploiting Intersubband Plasmons in Vertically Aligned Carbon Nanotubes for Near-Infrared Electrochromic Windows

Optically transparent materials with switchable near-infrared (NIR) transmissivity are of significant interest for energy-saving smart window technologies. To this end, we demonstrate that semitransparent films of vertically aligned carbon nanotubes (CNTs) incorporated into electrochemically gated devices exhibit NIR transmittance changes up to 47% and bistable optical states that are appealing for low-power, large-area operation. The tunable NIR electrochromic response is driven by a doping-induced intersubband plasmon (ISBP) absorption, an optical feature in CNTs that is selective to light polarized perpendicular to the CNT axis. Vertically aligned CNT films (as opposed to more conventional planar CNT mats) thus allow us to isolate and study the ISBP resonance changes with applied voltage, electrode material, and film thickness.

absorption↗
DOE OSTI2023

Uncertainty Aware Deep Learning for Fault Prediction Using Multivariate Time Series Signals

The superconducting radio-frequency cavities are a crucial component of the Continuous Electron Beam Accelerator Facility (CEBAF) at Jefferson Lab. When a cavity faults, beam delivery to experimental end users is disrupted. Prediction of cavity faults prior to onset is essential to reduce operation and maintenance costs. In this work, a parallel long short-term memory (LSTM)-convolution neural network (CNN)-based deep learning (DL) model is proposed to predict impending faults using pre-fault signals. Further, we introduce an uncertainty quantification approach using Monte Carlo dropout with the LSTM-CNN model to ascertain confidence in the prediction. The model was tested using multivariate time series signals from stable cavity operations and before faults. Initial results show that on the test dataset, the model can identify impending faults before their onset with an average 10-fold cross validation accuracy of 97.39% and a standard deviation of 0.12% using a 100-ms time window. It is also observed that the model performs better as the prediction time moves closer to the fault onset. For additional context, we compare the performance of the model with three machine-learning-based (ML) fault prediction models. Our proposed parallel LSTM-CNN-based DL method shows better performance than the ML-based methods.

Rahman, Md Monibor↗
DOE OSTI2025

The real-time data processing and acquisition system for Project 8 Phase II

In Phase II of the Project 8 neutrino mass experiment, electrons from the decays of tritium or 83 m Kr are detected via their ≈ 26GHz cyclotron radiation while contained within a circular waveguide. The signal from a given electron is characterized as a brief chirp, lasting ≲ 10ms and changing in frequency by ≲ 1MHz/ms. To detect these signals, the Project 8 collaboration developed a Data Acquisition (DAQ) system tailored to the signal properties. The DAQ is responsible for simultaneously selecting up to three 100MHz-wide frequency windows to study, detect, and trigger on likely signals from different electron kinetic energies, and for writing the relevant data to disk. We describe the Phase II DAQ system in detail and address how the system was used for data-taking operations.

Ashtari Esfahani, A. [Washington U., Seattle; Shar↗
DOE OSTI2020

A low temperature unitized regenerative fuel cell realizing 60% round trip efficiency and 10 000 cycles of durability for energy storage applications

Unitized regenerative fuel cells (URFC) convert electrical energy to and from chemical bonds in hydrogen. URFCs have the potential to provide economical means for efficient long-term, seasonal, energy storage and on-demand conversion back to electrical energy. We first optimize the catalyst layer for discrete electrolyzer and fuel cell and then configure the URFC. The goal is to identify a cost competitive configuration for URFCs, and demonstrate it in terms of upper limit of round trip efficiencies (RTEs). Two possible configurations of URFCs are compared via experiments and techno-economic analysis (TEA), which emphasize the advantages of the unconventional constant-electrode (CE) URFC over the traditional constant-gas (CG) configuration. We also study the stability via accelerated stress tests (ASTs) and demonstrate steady state operation in a daily cycle for day to night energy shifting. From the investigations, the optimum composition of the URFC anode catalyst layer is 90 at% Ir-black balanced by Pt-black for both CE and CG configurations. At 80 °C and 1 A cm -2 , the optimized CE URFC achieves 57% and 60% RTE with air and O 2 as the reductant gases, respectively. We then evaluated the differences in durability using an AST over 10k charge–discharge cycles; the results reveal that the wider potential window at the anode in CE (0.05–1.55 V) has minimal effect on catalyst layer stability compared to CG (0.55–1.55 V). Furthermore, there was no degradation up to the range of 2k–5k cycles; beyond that the fuel cell (discharge) performance degraded while the electrolyzer (charge) performance was still stable. Finally, the observations here indicate substantial potential to employ URFCs as efficient and cost-effective bidirectional energy-conversion devices within energy storage and utilization systems after appropriate technological and operational optimizations.

25 ENERGY STORAGE↗
DOE OSTI2026

ClassNMSW- a real-time classification approach for non-recycled municipal solid waste using hyperspectral imaging

Real-time classification of non-recycled municipal solid waste (NMSW) is essential for efficient valorization. This study introduces ClassNMSW, a comprehensive framework for classifying 22 NMSW subclasses under industrial constraints by using hyperspectral imaging (HSI). A primary innovation of this work is the development of a variance-controlled spectral extraction algorithm. Unlike traditional methods that rely on simple averaging, this approach systematically investigates the extent of pixel extraction to minimize the loss of critical chemical information while maximizing data reduction thus ensuring high spectral fidelity with low computational cost. The approach developed in this work integrates automated, computer-vision-based background removal, eliminating the need for the manual thresholding common in current literature. To resolve ambiguities among chemically similar subclasses, a tiered classification and multi-camera fusion strategy (NIR17 and NIR22) is implemented. Results demonstrate that ClassNMSW achieves an object-wise weighted accuracy of 98.70% for single-sensor configurations and 100% under sensor fusion. A novel rolling-window strategy satisfies desired end-to-end latency of <2 s, satisfying the strict deterministic requirements of high-speed industrial sorting environments. The ClassNMSW framework provides a scalable foundation for advancing circularity and resource recovery in large-scale waste valorization operations.

99 - GENERAL AND MISCELLANEOUS↗
DOE OSTI2025

The real-time data processing and acquisition system for Project 8 Phase II

In Phase II of the Project 8 neutrino mass experiment, electrons from the decays of tritium or 83m Kr are detected via their ≈26 GHz cyclotron radiation while contained within a circular waveguide. The signal from a given electron is characterized as a brief chirp, lasting ≲10 ms and changing in frequency by ≲1 MHz/ms. To detect these signals, the Project 8 collaboration developed a Data Acquisition (DAQ) system tailored to the signal properties. The DAQ is responsible for simultaneously selecting up to three 100 MHz-wide frequency windows to study, detect, and trigger on likely signals from different electron kinetic energies, and for writing the relevant data to disk. Finally, we describe the Phase II DAQ system in detail and address how the system was used for data-taking operations.

Data acquisition↗
DOE OSTI2026

torch-einshard v1.0

torch-einshard is a Python library for describing local and distributed PyTorch tensor computations with compact, einsum-like notation. Its expressions name logical axes, specify how they are sharded across a PyTorch DeviceMesh, and represent partial reductions. The library automatically performs contractions, permutations, reshaping, splitting, gathering, reduction, reduce-scatter, and repartitioning while preserving autograd. Additional features include sharding-aware FFTs, tensor rolls, halo exchange, sliding windows, 1D–3D convolutions, uneven-shard handling, parameter initialization and gradient management, and cost-based execution planning. It is designed for scientific machine learning and large-model workloads, including tensor-, sequence-, and spatial-parallel MLPs, attention, convolutions, and spectral operations. Compared with manually combining torch.einsum and distributed collectives, torch-einshard expresses both the mathematical operation and data placement in one readable formula. This reduces boilerplate and synchronization errors, keeps forward and backward communication consistent, and allows the library to select optimized collective strategies without changing model code.

Morozov, Dmitriy [Lawrence Berkeley National Labor↗
DOE OSTI2020

Large Volume Airborne Contamination Monitoring To Support Nuclear Processes' Deactivation and Decommissioning

Current D and D operations at Hanford have demonstrated a flaw in the current state of the art capability of defining airborne contamination boundaries - Airborne particulate emissions of Pu-239 from CM2H operations on the Hanford Pu Finishing Plant were detected well beyond areas controlled for airborne Pu, putting numerous workers at risk for radiological assimilations - Current air samplers and CAM systems were surveying insufficient volumes of air to accurately predict where respiratory protection was required - Hanford is located in a high radon environment, making air monitoring for alpha emitting actinides challenging in high alpha radon induced backgrounds. - Inexpensive HEPA based home and industrial air purifiers filter significantly higher volumes of air than commercially available continuous-air monitoring (CAM) systems. - Inexpensive models capable of filtered air volumes exceeding 1000 times that of a CAM. - Detector system can be built to detect x-rays generated from actinide decay. - Branching ratios of x-rays are 4 orders of magnitude more intense from Plutonium decay than its gamma emissions. - Detector system can be a low resolution systems, as x-ray region of interest is not terribly congested. - Detection systems evaluated will be designed based on slabs of NaI or pixelated NaI or CsI panels. - Simple graphical user interface software will be developed to operate the detection system. - Concept is to deploy some of these air-purifying systems in nonradiological areas around SRNL to confirm radon can be rejected with confidence. - Generate some filters contaminated with plutonium via electroplated Pu or lab generated simulated particles. - then deploy units in known airborne radiological environments to establish systems ability to accurately measure actinide based hot particles Lab analyses will follow up the system analyses to ensure hot particles were correctly identified. Analysis by Scintillation: - Application of scintillation media to the surface of the filters is being explored. - Evaluating slabs of ZnS(Ag), application of powdered ZnS(Cu) and spray on Perkin Elmer Enhance. - Scintillation events would then be digitized with a digital camera and quantified. - Comparing against sensitivity of a PMT or SiPD readout. - Currently evaluating a Thorlabs 8 Megapixel Monochrome Scientific CCD Camera, hermetically sealed cooled package with a wide angle lens. - Wide angle lens allows complete view of HEPA filter from 7 inches away. - Images taken of glow in the dark paint, as well as plutonium-induced fluorescence. - Plutonium was flamed mounted on a 1 inch diameter stainless steel planchet, covered with a layer of mylar and a section of scintillating ZnS obtained from Eljon. - Currently working on reducing signal-to-noise levels to boost sensitivity. - Dark box to hold contaminated or electroplated source covered filters under fabrication. Analysis by x-ray Spectroscopy: - Branching ratios for x-ray emissions from Plutonium isotopes are orders of magnitude more intense than gamma-ray emissions. - The nuclear databases are incomplete on the branching ratios of some of the isotopes. - Evaluating some custom built SrI2 x-ray spectrometers vs a conventional windowed NaI detector. - SrI2 detectors are carbon-composite-windowed 51 mm x 51 mm. - MCNP calculations to establish self absorbance of filter media on 17 keV x-ray. Radon Rejection: - Activated charcoal pre-filter acts as a Radon trap. - Adding a time lapse feature to camera to aid in radon rejection. SRS Plutonium Fuel Form D and D Operations: Currently deploying air sampler to D and D operation of Pu-238 facility at SRS to generate some field samples to analyze.

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗
DOE OSTI2020

Lyapunov growth in quantum spin chains

The Ising spin chain with longitudinal and transverse magnetic fields is often used in studies of quantum chaos, displaying both chaotic and integrable regions in its parameter space. However, even at a strongly chaotic point this model does not exhibit Lyapunov growth of the commutator squared of spin operators, as this observable saturates before exponential growth can manifest itself (even in situations where a spatial suppression factor makes the initial commutator small). In this work, we extend this model from the spin 1/2 Ising model to higher spins, demonstrate numerically that a window of exponential growth opens up for sufficiently large spin, and extract a quantity which corresponds to a notion of a Lyapunov exponent. In the classical infinite-spin limit, we identify and compute the appropriate classical analog of the commutator squared, and show that the corresponding exponent agrees with the infinite-spin limit extracted from the quantum spin chain.

75 CONDENSED MATTER PHYSICS, SUPERCONDUCTIVITY AND↗
DOE OSTI2025

Recyclable Design for Retaining High Solar Absorptivity of the Media in CSP

Efficient thermal energy storage is pivotal to lowering the levelized cost of electricity (LCOE) for Concentrating Solar Power (CSP) plants. In solid-particle systems, however, prolonged high-temperature service degrades particle solar absorptivity, eroding overall efficiency. This project demonstrates a hydrogen-assisted recovery process that reliably restores absorptivity to >90 %, offering a practical route to sustain long-term CSP performance. Bench-scale investigations mapped the reduction kinetics of optically faded particles across hydrogen concentrations, temperatures, and residence times. Coupling mass-spectrometric monitoring with machine-learning optimization minimized energy demand while maximizing absorptivity gain. The resulting process window—moderate hydrogen partial pressures, 15–30 min dwell times, and temperatures well below initial calcination levels—cuts energy consumption well below that of incumbent re-blackening methods. A prototype recovery reactor processed multiple 2 kg batches with repeatable outcomes, confirming scalability and operational robustness. Integrated techno-economic analysis indicates material and operating cost reductions exceeding 15 % relative to conventional particle replacement or chemical re-coating, translating directly into lower LCOE for next-generation CSP facilities. By uniting fundamental reaction-kinetics insight with pragmatic engineering, this work advances the solid-particle pathway, delivering a cost-effective, field-deployable solution to one of CSP’s key durability challenges and strengthening the commercial outlook for high-temperature renewable power.

14 SOLAR ENERGY↗
DOE OSTI2025

Exploring the Nature of Neutrinos with the Deep Underground Neutrino Experiment (DUNE)

The Deep Underground Neutrino Experiment (DUNE) is a next-generation experimental program designed to study the behaviour of neutrino oscillation. DUNE will utilize a neutrino beam originating at Fermilab, near Chicago, and will leverage a detector at Fermilab (Near Detector) and a detector 1300 km away in South Dakota (Far Detector), south of Saskatchewan. In the first phase of DUNE, its Far Detector will comprise of two 10,000 ton (fiducial) liquid argon (LAr) time-projection chamber (TPC) modules – powerful tracking calorimeter detectors – placed nearly a mile underground. With this large, sensitive, underground detector, DUNE aims to collect a high statistics and pure sample of neutrinos at the Far Detector. This setup also offers the potential to study non-beam physical processes via e.g. neutrinos produced in the atmosphere, supernova neutrino bursts, and/or solar neutrinos, etc. A second phase will aim to add more detector mass and expand the program. The Near Detector will consist of a LAr TPC module as well: critical to constraining systematic uncertainties in the oscillation analysis. However, this LAr TPC will have a novel design using a pixel-based readout instead of the traditional wire-based readout. This and the segmentation of the LAr TPC into multiple units are crucial in mitigating the high multiplicity of neutrino interactions expected in any readout window given its proximity to the beam. The Near Detector will feature additional components and capability beyond the LAr TPC, allowing one to deeply characterize the neutrino flux. Due to the complexity of this experimental program, several smaller-scale prototype detectors have been operating to test, validate, and improve both the technical designs and software for processing and analyzing events. By operating in charged particle test beams or neutrino beams, several of the prototypes are also capable of producing valuable results. Canadian institutions are involved in the realization of the DUNE through efforts with both the Near and Far Detectors and prototypes. DUNE is anticipated to begin operating near the end of this decade/the beginning of the next. This talk will focus on the overall DUNE program, for example its ultimate plans, status, and the efforts with prototypes.

Howard, Bruce [York U., Canada; Fermilab]↗
DOE OSTI2024

High-Power Targetry R&D for Next-Generation Accelerator Target Facilities

Beam-intercepting devices such as beam windows and particle-production targets are critical components of accelerator target facilities for High Energy Physics (HEP) experiments. The high-power, pulsed structure of the particle beams used for these experiments leads to thermal shock and high-cycle fatigue in addition to radiation damage resulting from the accumulated particle fluence. This can lead to degradation of the target system s mechanical and thermal properties; considerably reducing their lifetimes and presenting substantial challenges to reliable operation of multi-MW class facilities. Recently several major accelerator facilities have been forced to operate at reduced power levels due to target survivability concerns. Furthermore, at Fermilab it is planned to increase the neutrino production beam power up to 2.4 MW in coming years. Therefore, timely R&D on the irradiated behavior of target system materials is critical to efficient operation of accelerator facilities and full utilization of recent accelerator power upgrades for HEP research. This talk will begin with an overview of high-power targetry, and the significant challenges presented by beam power increases expected for future HEP experiments. We will then cover several past materials irradiation studies that have been completed by the High-Power Targetry R&D group at Fermilab and its collaborators on common accelerator and target materials such as graphite, beryllium, titanium, and tungsten. Finally, we will conclude with a discussion of two novel materials investigations under way within the group; high-entropy alloys for beam window applications, and electrospun nanofibers to serve as particle production targets.

43 PARTICLE ACCELERATORS↗
DOE OSTI2026

Estimation of the Required Dipole Corrector Magnetic Field for PIP-II Injection Based on Beam Studies

The Fermilab Booster will accept a 600 µs beam pulse from the new superconducting Linac for PIP-II operations. The Booster is a rapid-cycling synchrotron that uses a resonant magnet circuit ramping at 15 Hz. For PIP-II, the cycle rate will increase to 20 Hz, and the injection pulse length will expand from 40 µs to 600 µs due to the lower output current from the new Linac. Because the Booste main bending field follows a sinusoidal waveform, the magnetic field is not constant during the extended injection window. The longer pulse length and higher repetition rate modify the beam orbit and can lead to increased beam losses. The Booster contains 48 dipole-corrector packages distributed across its 24 periods. Each package includes horizontal and vertical dipoles, quadrupole, sextupole, skew-quadrupole, and skew-sextupole elements. By driving the dipole correctors with an appropriately shaped sinusoidal waveform during injection, we can compensate the changing main field and create an effectively flat bending field—referred to as flat injection. Over the past several years, machine studies have been performed to characterize the required correction fields and to determine the corresponding power-supply specifications needed for PIP-II operation. In this presentation, we will summarize the study results and discuss the estimated magnetic field requirements and power supply parameters for achieving flat injection in the Booster.

Seiya, K. [Fermilab] (ORCID:0000000250576943)↗
DOE OSTI2025

Probing Top-Quark–Electron Interactions at Future Colliders

Top quark interactions offer a window into possible new high scale physics and many models of new physics predict that the top quark interactions will deviate significantly from those predicted by the standard model. We present an analysis of the experimental restrictions on anomalous 4-fermion 𝑒 + ⁢𝑒 − $⁢𝑡\bar{⁢𝑡}$ operators that is accurate to next-to-leading order (NLO) in both the electroweak and QCD interactions within the standard model effective field theory framework. At NLO, there is sensitivity to an extended set of anomalous interactions beyond those probed at leading order. A comparison of current limits from electroweak precision observables, along with expected future limits from Drell-Yan and $⁢𝑡\bar{⁢𝑡}$𝑒 +⁢ 𝑒 − production at the high luminosity LHC, from deep inelastic scattering at the EIC, and from projected sensitivities at the future FCC-ee and CEPC machines demonstrates that each of these programs extends the precision understanding of the interactions of top quarks.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2024

PV Generation and Load Forecasting for Adjuntas PR Community Microgrids

Existing frameworks to forecast time-series photovoltaic (PV) output power and consumer load for microgrid operations and controls assume a near-continuous availability of real-time input features from the field assets such as PV inverters, energy meters, and weather station. These incoming data points are used to periodically retrain models and update forecast snapshots over a moving horizon window, be it one hour-ahead, one-day ahead, or one-week ahead. However, such frameworks are not resilient to disruptions in data availability caused by losses in communications between the field sensors and data loggers. Hence, there is a need for programs that assume no availability of real-time microgrid asset data and still make reliable forecasts that can be used for decision-making. Such programs would be apt to function in extreme weather events such as hurricanes and would use lightweight recursive time-series models to independently forecast solar irradiance and ambient temperature, then compute PV power from those forecasts, as well as independently forecast consumer load. The codebase performs forecasting for the scenario of when the microgrid does not have a reliable access to forecasts or real-time observations of solar irradiance (I) and ambient temperature (AT) and load (Load) to be able to adequately forecast, in real-time, the PV power production or a business' load. In this case, using historical values of PV power and load, a univariate forecasting of generation and consumption are respectively made. The use-case in particular has two sub-scenarios: one, a normal 7-day ahead forecast where the unavailability of real-time data is assumed due to infrastructure issues such as loss of communication or sensor maintenance or service downtimes. Whereas a hurricane-caused unavailability of real-time data requires a second model trained specifically on historical hurricane days to be able to capture the extreme day behavior of generation in particular, and load if applicable. A gradient boosted regression tree comprises an ensemble of additive models that map between the input of historical values (be it irradiance, temperature, or load) and their corresponding output forecasts of a given horizon such that the individual learner predictions are summed up over the total number of such learners in the ensemble to produce an aggregate forecast. A weighting mechanism is applied to the training data in each iteration, where actual and forecast values are compared to penalize incorrect forecasts by increasing the weight and reducing it to reward correct forecasts. The code's benefits are that it: (a) accounts for a contingency where communication loss renders newly measured real-time data unavailable for model tuning and snapshot updates; (b) presents blind forecasting that recursively determines the next time-step value in a horizon using the forecast of the same attribute from a prior step; and (c) employs lightweight models that, once trained, can reliably generalize for different horizons, which make them suitable for enhancing the resilience of field microgrids prone to extreme events that encounter disruptions to data availability.

Sundararajan, Aditya [Oak Ridge National Laborator↗