Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Data models”

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 1,621 records · Page 90

Robust Predictive Design of Field Measurements for Evapotranspiration Barriers Using Universal Multiple linear Regression

Surface barriers are normally installed to reduce downward water movement into contaminated zones. Specifically, evapotranspiration (ET) barriers are used to store water and release it, via ET, before it can percolate into an underlying waste zone. To assess the effectiveness of a surface barrier, we used an existing data set, model-simulated data, and a dimensionality reduction approach called universal multiple linear regression (uMLR) to optimize the required number of sensors in a 2-m thick surface barrier. To understand the usefulness of implementing predictive uMLR to accommodate multiple monitoring objectives, we compare several network designs, selected based on down-sampling of existing data, with a recommended sensor design based on model simulations performed without consideration of existing data. We further added consideration of “fuzzy” design, which allows more practical guidelines for field implementation of uMLR. We discovered that uMLR, combined with robust decision-making, provides a simple, flexible, and high-quality network design for monitoring the total water stored in a surface barrier across multiple uncertain conditions.

58 GEOSCIENCES↗

Data Analysis and Model Validation (2013 CEF Run - Phase 1)

Phase 1 of the 2013 Cold cap Evaluation Furnace (CEF) test was completed on June 3, 2013 after a 5-day round-the-clock feeding and pouring operation. The main goal of the test was to characterize the CEF offgas produced from a nitric-formic acid flowsheet feed and confirm whether the CEF platform is capable of producing scalable off-gas data necessary for the revision of the DWPF melter off-gas flammability model; the revised model will be used to define new safety controls on the key operating parameters for the nitric-glycolic acid flowsheet feeds including total organic carbon (TOC). Whether the CEF off-gas data were scalable for the purpose of predicting the potential flammability of the DWPF melter exhaust was determined by comparing the predicted H 2 and CO concentrations using the current DWPF melter off-gas flammability model to those measured during Phase 1; data were deemed scalable if the calculated fractional conversions of TOC-to-H 2 and TOC-to-CO at varying melter vapor space temperatures were found to trend and further bound the respective measured data with some margin of safety. Being scalable thus means that for a given feed chemistry the instantaneous flow rates of H 2 and CO in the DWPF melter exhaust can be estimated with some degree of conservatism by multiplying those of the respective gases from a pilot-scale melter by the feed rate ratio. This report documents the results of the Phase 1 data analysis and the necessary calculations performed to determine the scalability of the CEF off-gas data.

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

Carbon Organisms Rhizosphere and Protection in Soil Environment model script and input data for soil moisture-respiration responses in tropical forests

Objectives: Climatic drying is predicted for many tropical forests, yet models remain poorly parameterized for tropical forests, hampering predictions of forest-climate feedbacks. We applied an integrated model–experiment approach, parameterizing an ecosystem model Carbon Organisms Rhizosphere and Protection in the Soil Environment (CORPSE) with tropical forest observational data, and comparing model predictions with a field drying manipulation. We hypothesized that drying would suppress soil CO2 fluxes (i.e., respiration) in already-drier tropical forests, but increases CO2 fluxes in wetter tropical forests by alleviating anaerobiosis. We measured soil CO2 fluxes, soil moisture, soil temperature, and forest floor biomass during wet-dry cycles (2015 – 2022) in four Panamanian forests that vary in rainfall and soil fertility. We used the field data to parameterize and run tests in the model.Results: Measured CO2 fluxes declined in the dry season and peaked in the early wet season ahead of peak soil moisture, resulting in a lower soil moisture optimum for respiration than previously modeled. We used this data to parameterize the model, which then predicted increased soil CO2 fluxes in wetter and fertile forests with drying, and decreased fluxes in drier, infertile forests. In contrast to model predictions, a chronic throughfall exclusion experiment in the forests initially suppressed soil CO2 fluxes across forests, with sustained suppression after four years in the wettest forest only (-28 ± 4% during the dry season), but elevated soil CO2 fluxes in a fertile forest after four years (+75 ± 28% during the late wet season), as predicted by the model. The unexpected negative drying effect in the wettest, most infertile forest could have resulted from reduced vertical flushing of nutrients into soils. Including hydro-nutrient interactions in ecosystem models could improve predictions of tropical forest-climate feedbacks (results presented in Cusack et al. 2023). Datasets included: Code files:CORPSE_array.py: Defines the equations of the CORPSE modelCORPSE_solvers: Functions for running the CORPSE model using either iterative or ordinary differential equation (ODE) solversrun_Panama_sims.py: Read in datasets and run the model simulations for this studyInput data:PanamaGradientEcosystemChem_BT_CPools_20152016CO2_DC_20190615.xlsx: Plot characteristics used in running model simulationsLiCor compiled surface flux only to 2020_03 DC_20200825.xlsx: Surface gas exchange fluxes used in model-data comparisonsPARCHED litterfall data for Ben Sulman LD 20200902.xlsx: Litterfall data used to drive model simulationsInitialization data:state_500y_20190823.csv: Initial state of model pools based on previous spinup runsOutput data:Outputs/prev_moisture_response.csv: Simulations of multiple sites using original model moisture response function.Outputs/updated_moisture_response.csv: Simulations of multiple sites using updated model moisture response function.Outputs/dry15_prev_moisture_response.csv: Simulations with soil moisture reduced by 15%, using original moisture response function.Outputs/dry15_updated_moisture_response.csv: Simulations with soil moisture reduced by 15%, using updated moisture response function.Outputs/dry30_prev_moisture_response.csv: Simulations with soil moisture reduced by 30%, using original moisture response function.Outputs/dry30_updated_moisture_response.csv: Simulations with soil moisture reduced by 30%, using updated moisture response function.Outputs/latestart_prev_moisture_response.csv: Simulations with extended dry season, using original moisture response function.Outputs/latestart_updated_moisture_response.csv: Simulations with extended dry season, using updated moisture response function.Outputs/[site name]_oneyear.csv: One-year simulation for each site in expanded site list using original moisture response function.Outputs/[site name]_oneyear_dried.csv: One-year simulation for each site in expanded site list using original moisture response function, with soil moisture reduced by 25%.Outputs/[site name]_oneyear_updated_moisture_response.csv: One-year simulation for each site in expanded site list using updated moisture response function.Outputs/[site name]_oneyear_updated_moisture_response_dried.csv: One-year simulation for each site in expanded site list using updated moisture response function, with soil moisture reduced by 25%.Field plot location data:There is also a .kml file that includes coordinates for all 32 plots included in the study of four forests (n = 4 throughfall reduction and n = 4 control plots per site).

54 ENVIRONMENTAL SCIENCES↗

Quantification of terrestrial ecosystem carbon dynamics in the conterminous United States combining a process-based biogeochemical model and MODIS and AmeriFlux data

Satellite remote sensing provides continuous temporal and spatial information of terrestrial ecosystems. Using these remote sensing data and eddy flux measurements and biogeochemical models, such as the Terrestrial Ecosystem Model (TEM), should provide a more adequate quantification of carbon dynamics of terrestrial ecosystems. Here we use Moderate Resolution Imaging Spectroradiometer (MODIS) Enhanced Vegetation Index (EVI), Land Surface Water Index (LSWI) and carbon flux data of AmeriFlux to conduct such a study. First we modify the gross primary production (GPP) modeling in TEM by incorporating EVI and LSWI to account for the effects of the changes of canopy photosynthetic capacity, phenology and water stress. Second, we parameterize and verify the new version of TEM with eddy flux data. We then apply the model to the conterminous United States over the period 2000–2005 at a 0.05° × 0.05° spatial resolution. We find that the new version of TEM made improvement over the previous version and generally captured the expected temporal and spatial patterns of regional carbon dynamics. We estimate that regional GPP is between 7.02 and 7.78 PgC yr -1 and net primary production (NPP) ranges from 3.81 to 4.38 Pg Cyr -1 and net ecosystem production (NEP) varies within 0.08– 0.73 PgC yr -1 over the period 2000–2005 for the conterminous United States. The uncertainty due to parameterization is 0.34, 0.65 and 0.18 PgC yr -1 for the regional estimates of GPP, NPP and NEP, respectively. The effects of extreme climate and disturbances such as severe drought in 2002 and destructive Hurricane Katrina in 2005 were captured by the model. Lastly, our study provides a new independent and more adequate measure of carbon fluxes for the conterminous United States, which will benefit studies of carbon-climate feedback and facilitate policy-making of carbon management and climate.

54 ENVIRONMENTAL SCIENCES↗

AI-Based EMT Dynamic Model of PV Systems

Several electromagnetic transient (EMT) dynamic modeling methods are available to model systems like photovoltaic (PV) plants, wind power plants, variable-speed drives, among others. The methods include: (a) physics-based models and (b) data-driven models. The physics-based dynamic models may include high-fidelity switched system model and average-value model that both require the control algorithms included in the models. However, manufacturers typically prefer to provide black-box models to avoid disclosing proprietary. One of the solutions to prevent disclosing control algorithms is the use of data-driven dynamic EMT models of PV systems. In this paper, data-driven dynamic EMT model based on artificial intelligence (AI) algorithms are presented. The AI algorithms evaluated include convolutional neural networks, recurrent neural networks, and nonlinear auto-regressive exogenous model. Automation in generating data and training these models is also discussed in this paper. The results generated by the best AI algorithms have been observed to be greater than 95 % accurate.

Debnath, Suman↗

Data Assimilation Tools for CO2 Reservoir Model Development – A Review of Key Data Types, Analyses, and Selected Software

Pacific Northwest National Laboratory (PNNL) has embarked on an initiative to develop world-class capabilities for performing experimental and computational analyses associated with geologic sequestration of carbon dioxide. The ultimate goal of this initiative is to provide science-based solutions for helping to mitigate the adverse effects of greenhouse gas emissions. This Laboratory-Directed Research and Development (LDRD) initiative currently has two primary focus areas—advanced experimental methods and computational analysis. The experimental methods focus area involves the development of new experimental capabilities, supported in part by the U.S. Department of Energy’s (DOE) Environmental Molecular Science Laboratory (EMSL) housed at PNNL, for quantifying mineral reaction kinetics with CO2 under high temperature and pressure (supercritical) conditions. The computational analysis focus area involves numerical simulation of coupled, multi-scale processes associated with CO2 sequestration in geologic media, and the development of software to facilitate building and parameterizing conceptual and numerical models of subsurface reservoirs that represent geologic repositories for injected CO2. This report describes work in support of the computational analysis focus area. The computational analysis focus area currently consists of several collaborative research projects. These are all geared towards the development and application of conceptual and numerical models for geologic sequestration of CO2. The software being developed for this focus area is referred to as the Geologic Sequestration Software Suite or GS3. A wiki-based software framework is being developed to support GS3. This report summarizes work performed in FY09 on one of the LDRD projects in the computational analysis focus area. The title of this project is Data Assimilation Tools for CO2 Reservoir Model Development. Some key objectives of this project in FY09 were to assess the current state-of-the-art in reservoir model development, the data types and analyses that need to be performed in order to develop and parameterize credible and robust reservoir simulation models, and to review existing software that is applicable to these analyses. This report describes this effort and highlights areas in which additional software development, wiki application extensions, or related GS3 infrastructure development may be warranted.

54 ENVIRONMENTAL SCIENCES↗

Machine learning-based surrogate models and transfer learning for derivative free optimization of HT-PEM fuel cells

Widespread adoption of high-temperature polymer electrolyte membrane electrochemical systems, such as fuel cells (HT-PEMFCs), requires models and computational tools for accurate optimization and guiding new materials for enhancing performance and durability. In this contribution, knowledge-based modelling and data-driven modelling are combined using Few-Shot Learning and implementing an Automated Machine Learning framework for the generation of Machine Learning-based surrogate models. Applicability of the resulting model for derivative-free optimization is demonstrated. Additionally, a way of considering extrapolation in the optimization task is presented. Results show that although extrapolation is needed to achieve better solutions during optimization, it can be monitored and managed. As a result, tuning the electrode ionomer binder's properties, such as ionic conductivity, in the fuel cell represents a promising pathway for improving HT-PEMFC performance.

08 HYDROGEN↗

Using Large Language Models to help customers monitor global threat data

Large Language Models have proven adept at answering general knowledge questions. To make these generative AI tools useful to our mission customers for monitoring global threats, the data sciences team at Sandia is utilizing retrieval augmented generation (RAG) techniques to customize these models with local data. The local data we use consists of data such as research articles and patent abstracts that we've collected over the last several years using automated pipelines.

Herzer, John Andrew [Sandia National Laboratories ↗

Baryon acoustic oscillation, Hubble parameter, and angular size measurement constraints on the Hubble constant, dark energy dynamics, and spatial curvature

ABSTRACT In this paper, we use all available baryon acoustic oscillation, Hubble parameter, and quasar angular size data to constrain six dark energy cosmological models, both spatially flat and non-flat. Depending on the model and data combination considered, these data mildly favour closed spatial hypersurfaces (by as much as 1.7σ) and dark energy dynamics (up to a little over 2σ) over a cosmological constant Λ. The data also favour, at 1.8σ to 3.4σ, depending on the model and data combination, a lower Hubble constant than what is measured from the local expansion rate.

79 ASTRONOMY AND ASTROPHYSICS↗

U.S. emissions of HFC-134a derived for 2008-2012 from an extensive flask-air sampling network: US emissions of HFC-134a

U.S. national and regional emissions of HFC-134a are derived for 2008-2012 based on atmospheric observations from ground and aircraft sites across the U.S. and a newly developed regional inverse model. Synthetic data experiments were first conducted to optimize the model assimilation design and to assess model-data mismatch errors and prior flux error covariances computed using a maximumlikelihood estimation technique. The synthetic data experiments also tested the sensitivity of derived national and regional emissions to a range of assumed prior emissions, with the goal of designing a system that was minimally reliant on the prior. We then explored the influence of additional sources of error in inversions with actual observations, such as those associated with background mole fractions and transport uncertainties. Estimated emissions of HFC-134a range from 52 to 61 Gg yr -1 for the contiguous U.S. during 2008-2012 for inversions using air transport from Hybrid Single-Particle Lagrangian Integrated Trajectory (HYSPLIT) model driven by the 12km resolution meteorogical data from North American Mesoscale Forecast System (NAM12) and all tested combinations of prior emissions and background mole fractions. Estimated emissions for 2008-2010 were 20% lower when specifying alternative transport from Stochastic Time-Inverted Lagrangian Transport (STILT) model driven by the Weather Research and Forecasting (WRF) meteorology. Our estimates (for HYSPLIT-NAM12) are consistent with annual emissions reported by U.S. Environmental Protection Agency for the full study interval. The results suggest a 10-20% drop in U.S. national HFC-134a emission in 2009 coincident with a reduction in transportation-related fossil fuel CO 2 emissions, perhaps related to the economic recession. All inversions show seasonal variation in national HFC-134a emissions in all years, with summer emissions greater than winter emissions by 20-50%.

54 ENVIRONMENTAL SCIENCES↗

Data-driven reduced-order models for port-Hamiltonian systems with operator inference

Hamiltonian operator inference has been developed in Sharma et al. (2022) to learn structure-preserving reduced-order models (ROMs) for Hamiltonian systems. The method constructs a low-dimensional model using only data and knowledge of the functional form of the Hamiltonian. The resulting ROMs preserve the intrinsic structure of the system, ensuring that the mechanical and physical properties of the system are maintained. In this work, we extend this approach to port-Hamiltonian systems, which generalize Hamiltonian systems by including energy dissipation, external input, and output. Based on snapshots of the system’s state and output, together with the information about the functional form of the Hamiltonian, reduced operators are inferred through optimization and are then used to construct data-driven ROMs. To further alleviate the complexity of evaluating nonlinear terms in the ROMs, a hyper-reduction method via discrete empirical interpolation is applied. Accordingly, we derive error estimates for the ROM approximations of the state and output. Lastly, we demonstrate the structure preservation, as well as the accuracy of the proposed port-Hamiltonian operator inference framework, through numerical experiments on a linear mass–spring-damper problem and a nonlinear Toda lattice problem.

97 MATHEMATICS AND COMPUTING↗

PREDICTING CME EJECTA AND SHEATH FRONT ARRIVAL AT L1 WITH A DATA-CONSTRAINED PHYSICAL MODEL

We present a method for predicting the arrival of a coronal mass ejection (CME) flux rope in situ, as well as the sheath of solar wind plasma accumulated ahead of the driver. For faster CMEs, the front of this sheath will be a shock. The method is based upon geometrical separate measurement of the CME ejecta and sheath. These measurements are used to constrain a drag-based model, improved by including both a height dependence and accurate de-projected velocities. We also constrain the geometry of the model to determine the error introduced as a function of the deviation of the CME nose from the Sun–Earth line. The CME standoff-distance in the heliosphere fit is also calculated, fit, and combined with the ejecta model to determine sheath arrival. Combining these factors allows us to create predictions for both fronts at the L1 point and compare them against observations. We demonstrate an ability to predict the sheath arrival with an average error of under 3.5 hr, with an rms error of about 1.58 hr. For the ejecta the error is less than 1.5 hr, with an rms error within 0.76 hr. We also discuss the physical implications of our model for CME expansion and density evolution. We show the power of our method with ideal data and demonstrate the practical implications of having a permanent L5 observer with space weather forecasting capabilities, while also discussing the limitations of the method that will have to be addressed in order to create a real-time forecasting tool.

79 ASTRONOMY AND ASTROPHYSICS↗

Modeling aerosol bolus inhalations in the human lung with the multiple path particle deposition model: Comparison with experimental data

Existing one-dimensional (1D) models of aerosol dosimetry often ignore mixing mechanisms of inhaled aerosols during their transport in the lung. This mixing or aerosol dispersion results from different physical mechanisms in different regions of the lung. It is a higher order effect, which cannot be directly captured in 1D modeling approaches, and thus is sometimes modeled as a diffusive process. Here, in this study, we improved our recently developed alveolar mixing module incorporated in the multiple path particle dosimetry model (MPPD) to account for flow irreversibility and particle trapping in the alveolar spaces, as well as mixing occurring in the tracheobronchial region. This new version of MPPD was coupled with CFPD-based predictions of aerosol bolus dispersion in the oral airway. The model was used to predict the deposition, dispersion, and mode shift of aerosol bolus inhaled at different penetration depths within the lung for breathing patterns and particle size matching those used in a previous experimental study (Darquenne et al., 2016). Even though a quite simplified approach was used, the computations appear to describe subject-specific and test-specific experimental data reasonably well. The proposed combined dispersion-deposition model can be a useful tool for targeted drug delivery and also for exposure health risk assessment.

MPPD↗

Loose Belt Fault Detection and Virtual Flow Meter Development Using Identified Data-driven Energy Model for Fan Systems

An energy model that correlates fan airflow, head, speed, and system power input is essential to detect device faults and optimize control strategies in fan systems. Since the application of variable-frequency drives (VFDs) makes the motor-efficiency data published by manufacturers inapplicable for VFD–motor–fan systems, the fan efficiency and drive (belt–motor–VFD) efficiency must be identified for each individual system to obtain accurate energy models. The objectives of this paper are to identify an energy model of existing VFD–motor–fan systems using available experimental data and demonstrate its applications in loose belt fault detection and virtual airflow meter development for optimal control. First, an approach is developed to identify the fan head, fan efficiency, and drive-efficiency curves using available fan head, speed, and system power input as well as temporarily measured airflow rate without measuring shaft power. Then, the energy model is identified for an existing VFD–motor–fan system. Finally, the identified model is applied to detect the slipped belt faults and develop the virtual airflow meter. The experiment results reveal that the developed approach can effectively obtain the energy model of VFD–motor–fan systems and the model can be applied to effectively detect slipped belt faults and accurately calculate the fan airflow rate.

42 ENGINEERING↗

Global tuning of hadronic interaction models with accelerator-based and astroparticle data

In high-energy and astroparticle physics, event generators play an essential role, even in the simplest data analyses. As analysis techniques become more sophisticated, e.g. based on deep neural networks, their correct description of the observed event characteristics becomes even more important. Physical processes occurring in hadronic collisions are simulated within a Monte Carlo framework. A major challenge is the modeling of hadron dynamics at low momentum transfer, which includes the initial and final phases of every hadronic collision. QCD-inspired phenomenological models used for these phases cannot guarantee completeness or correctness over the full phase space. These models usually include parameters which must be tuned to suitable experimental data. Until now, event generators have been developed and tuned mainly on the basis of data from high-energy physics experiments at accelerators. The wealth of data available from the latest generation of astroparticle experiments has not yet been fully exploited, and in many cases is not satisfactorily described. Both kinds of data sets are complementary as astroparticle experiments provide sensitivity especially to hadrons produced nearly parallel to the collision axis and cover center-of-mass energies up to several hundred TeV, well beyond those reached at colliders so far. In this report, we provide an overview of state-of-the-art event generators and their tuning, including the most relevant inputs from high-energy accelerator and astroparticle experiments. We present a road map that shows, for the first time, how the unified tuning of event generators with accelerator-based and astroparticle data can be performed.

Albrecht, J. [Ruhr U., Bochum, RAPP Ctr.; Ruhr U.,↗

A field guide for aging passerine nestlings using growth data and predictive modeling

Abstract Background Accurate nestling age is valuable for studies on nesting strategies, productivity, and impacts on reproductive success. Most aging guides consist of descriptions and photographs that are time consuming to read and subjective to interpret. The Western Bluebird ( Sialia mexicana ) is a secondary cavity-nesting passerine that nests in coniferous and open deciduous forests. Nest box programs for cavity-nesting species have provided suitable nesting locations and opportunities for data collection on nestling growth and development. Methods We developed models for predicting the age of Western Bluebird nestlings from morphometric measurements using model training and validation. These were developed for mass, tarsus, and two different culmen measurements. Results Our models were accurate to within less than a day, and each model worked best for a specific age range. The mass and tarsus models can be used to estimate the ages of Western Bluebird nestlings 0–10 days old and were accurate to within 0.5 days for mass and 0.7 days for tarsus. The culmen models can be used to estimate ages of nestlings 0–15 days old and were also accurate to within less than a day. The daily mean, minimum, and maximum values of each morphometric measurement are provided and can be used in the field for accurate nestling age estimations in real time. Conclusions The model training and validation procedures used here demonstrate that this method can create aging models that are highly accurate. The methods can be applied to any passerine species provided sufficient nestling morphometric data are available.

59 BASIC BIOLOGICAL SCIENCES↗

SU-E-T-73: Commissioning of a Treatment Planning System for Proton Spot Scanning

Purpose: A treatment planning system (TPS) was commissioned for clinical use with a fixed beam line proton delivery system. An outline of the data collection, modeling, and verification is provided. Methods: Beam data modeling for proton spot scanning in CMS Xio TPS requires the following measurements: (i) integral depth dose curves (IDDCs); (ii) absolute dose calibration; and (iii) beam spot characteristics. The IDDCs for 18 proton energies were measured using an integrating detector in a single spot field in a water phantom. Absolute scaling of the IDDCs were performed based on ion chamber measurements in mono-energetic 10×10 cm{sup 2} fields in water. Beam spot shapes were measured in air using a flat panel scintillator detector at multiple planes. For beam model verification, more than 45 uniform dose phantom and patient plans were generated. These plans were used to measure range, point dose, and longitudinal and lateral profiles. Tolerances employed for verification are: point dose and longitudinal profiles, ±2%; range, ±1 mm; FWHM for lateral profiles, ±2 mm; and patient plan dose distribution, gamma index of >90% at 3%/3 mm criteria. Results: More than 97% of the point dose measurements out of 115 were within +/-2% with maximum deviation of 3%. 98% of the ranges measured were within 1 mm with maximum deviation of 1.4mm. The normalized depth doses were within 2% at all depths. The maximum error in FWHM of lateral profiles was found to be less than 2mm. For 5 patient plans representing different anatomic sites, a total of 38 planes for 12 beams were analyzed for gamma index with average value of 99% and minimum of 94%. Conclusions: The planning system is successfully commissioned and can be safely deployed for clinical use. Measurements of IDDCs on user beam are highly recommended instead of using standard beam IDDCs.

60 APPLIED LIFE SCIENCES↗

Data-based Aggregate Model of Refrigerators for Electric Power Grid Services

Electrical power system stability under high penetration of intermittent renewable energy generation, such as wind and solar energy, has become a high concern. It is unpredictable and hard to control due to little inertial behavior characteristic to these generation resources. To help mitigate possibly undesirable frequency excursions, demand-side frequency control presents an emerging solution. By observing the frequency deviation, devices could switch operating modes for a short time to rebalance generation. To analyze device level control algorithms, aggregate models for their behavior are needed. Refrigerators are household devices that could help with load modulation. This paper uses real measured data to analyze the refrigerator behavior, learn about its operating mode and power consumption ranges. A Markov chain process is proposed to model the transition of a refrigerator population from one operating mode to another, such that the overall load is modulated according to the control input. The Markov-chain model can be used to design control so that the overall load is modulated according to the control input.

Markov chain model, demand response, refrigerator,↗