Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Optimization Under Uncertainty”

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 1,549 records · Page 86

Deep Reinforcement Learning for Distribution System Operations: A Tutorial and Survey

Here, the rapid evolution of modern electric power distribution systems into complex networks of interconnected active devices, distributed generation (DG), and storage poses increasing difficulties for system operators. The large-scale integration of distributed energy resources (DERs) and the rapid exchange of measurement data via communication networks present major opportunities for advancing grid operations but also introduce greater uncertainty, higher data dimensionality, more complex network and device models, and challenging control and optimization problems. Deep reinforcement learning (DRL) algorithms are promising in addressing these challenges. However, they have not been effectively adapted for power systems applications, requiring extensive customization for implementation and evaluation. This has resulted in reproducibility challenges and a steep learning curve for researchers new to applying DRL algorithms to the power systems domain. To bridge these gaps, this tutorial aims to serve as a valuable resource for researchers interested in exploring learning-based algorithms to operate active power distribution networks. Specifically, this work presents a generalized process for translating sequential decision-making problems in power distribution systems into Markov decision process (MDP) formulations, illustrated through concrete grid service examples. Additionally, we introduce a simple environment design strategy to develop and evaluate example DRL algorithms for distribution system applications, complete with an included code repository to guide users through environment construction.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Multi-Agent Hierarchical Deep Reinforcement Learning for HVAC Control With Flexible DERs

As electricity consumption in commercial and residential buildings continues to rise, reducing energy costs presents an increasing challenge. Heating, ventilating, and air-conditioning (HVAC) systems, which typically account for 40%-50% of a building's energy use, are prime targets for energy savings. Intelligent control of HVAC temperature through the exploitation of HVAC load flexibility brings significant potential to reduce energy consumption and electricity expenses. The nonlinear models of HVAC systems challenge traditional control methods, while the uncertainty introduced by HVAC load flexibility complicates distributed energy resource (DER) management using conventional optimal dispatch techniques. In response to these challenges, we propose a hierarchical multi-agent deep reinforcement learning (DRL) approach. The lower-level agents focus on balancing comfort and energy conservation, while the upper-level DRL agents optimize the use of DERs to reduce peak demand based on the control outcomes of the HVAC by the lower-level agents. Here, in the upper-level agents, we incorporate a multi-agent structure based on ensemble learning, which acts based on historical and current data without relying on precise load forecasting to address the delayed rewarding issue in DRL. This allows for the effective reduction of energy costs. The proposed method is tested using a real-world microgrid comprising 413 buildings in Southern California, and the results demonstrate that our approach can significantly reduce overall electricity bills while ensuring the comfort of consumers and residents.

24 POWER TRANSMISSION AND DISTRIBUTION↗

TH-C-19A-10: Systematic Evaluation of Photodetectors Performances for Plastic Scintillation Dosimetry

Purpose: To assess and compare the performance of different photodetectors likely to be used in a plastic scintillation detector (PSD). Methods: The PSD consists of a 1 mm diameter, 10 mm long plastic scintillation fiber (BCF-60) which is optically coupled to a clear 10 m long optical fiber of the same diameter. A light-tight plastic sheath covers both fibers and the scintillator end is sealed. The clear fiber end is connected to one of the following six studied photodetectors: two polychromatic cameras (one with an optical lens and one with a fiber optic taper replacing the lens); a monochromatic camera with the same optical lens; a PIN photodiode; an avalanche photodiode (APD); and a photomultiplier tube (PMT). Each PSD is exposed to both low energy beams (120, 180, and 220 kVp) from an orthovoltage unit, and high energy beams (6 MV and 23 MV) from a linear accelerator. Various dose rates are explored to identify the photodetectors operating ranges and accuracy. Results: For all photodetectors, the relative uncertainty remains under 5 % for dose rates over 3 mGy/s. The taper camera collects four times more signal than the optical lens camera, although its standard deviation is higher since it could not be cooled. The PIN, APD and PMT have higher sensitivity, suitable for low dose rate and out-of-field dose monitoring. PMT's relative uncertainty remains under 1 % at the lowest dose rate achievable (50 μGy/s), suggesting optimal use for live dosimetry. Conclusion: A set of 6 photodetectors have been studied over a broad dose rate range at various energies. For dose rate above 3 mGy/s, the PIN diode is the most effective photodetector in term of performance/cost ratio. For lower dose rate, such as those seen in interventional radiology, PMTs are the optimal choice. FQRNT Doctoral Research Scholarship.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

SU-E-T-457: Impact of Interfractional Variations On Anterior Vs. Lateral-Field Proton Therapy of Prostate Cancer

Purpose: To investigate the effects of interfractional anatomy and setup variations on plans with anterior-oblique vs. lateral beams for prostate cancer pencil beam scanning (PBS) and passive scattered (PS) proton therapy. Methods: Six patients with low/intermediate risk prostate cancer treated with PS proton therapy at our institution were selected. All patients underwent weekly verification CT scans. Implanted fiducials were used for localization, and endorectal balloons for prostate immobilization. New PBS plans with lateral beams, as well as PBS and PS plans with anterior-oblique beams (±35 deg) were created. PBS plans used two different spot sizes: ∼10mm (large) and ∼5mm (medium) sigma at 25cm range and optimized as single-field-uniform-dose with ∼8% non-uniformity. No range uncertainty margins were applied in PBS plans to maximize rectal sparing. Field-specific apertures were used when planning with large spots to sharpen the penumbrae. The planned dose was recomputed on each weekly CT with fiducials aligned to the simulation CT, scaled and accumulated via deformable image registration. Results: The dose volume analysis showed that although difference between planned and accumulated dose remains negligible for plans with conventional lateral beams using both PS and PBS, this is not the case for plans with anterior beams. The target coverage in anterior plans was largely degraded due to the variations in the beam path length and the absence of range margins. The average prostate D95 was reduced by 7.5/15.9% (using PS/PBS) after accumulation for anterior plans, compared with 0/0.4% for lateral plans. The average mean dose in organs-at-risk decreased by 1% for lateral and 2% for anterior plans, similarly for PS and PBS. Spot size did not affect the dose changes. Conclusion: Prostate plans using anterior beams may undergo clinically relevant interfractional dose degradation. Corrective strategies guided by in-vivo range measurements should be studied before clinical application of this technique.

60 APPLIED LIFE SCIENCES↗

Enhancing Gaussian Process Surrogates for Optimization and Posterior Approximation via Random Exploration

This paper proposes novel noise-free Bayesian optimization strategies that rely on a random exploration step to enhance the accuracy of Gaussian process surrogate models. The new algorithms retain the ease of implementation of the classical GP-UCB algorithm, but the additional random exploration step accelerates their convergence, nearly achieving the optimal convergence rate. Furthermore, to facilitate Bayesian inference with intractable likelihoods, we propose to utilize optimization iterates for maximum a posteriori estimation to build a Gaussian process surrogate model for the unnormalized log-posterior density. We provide bounds for the Hellinger distance between the true and the approximate posterior distributions in terms of the number of design points. We demonstrate the effectiveness of our Bayesian optimization algorithms in nonconvex benchmark objective functions, in a machine learning hyperparameter tuning problem, and in a black-box engineering design problem. The effectiveness of our posterior approximation approach is demonstrated in two Bayesian inference problems for parameters of dynamical systems.

Bayesian inference↗

Historic climate, cosmogenic 10Be, denudation-rate, and geospatial datasets from the Pikes Peak region, Colorado, USA

This data package contains geographic information system (GIS) layers and tabular datasets associated with the study of elevation-dependent denudation rates on Pikes Peak in the Front Range of the Rocky Mountains, Colorado, USA. The package includes GIS layers used to produce the study-area map, including sample locations, sample watershed boundaries, the Pikes Peak batholith, Pleistocene glacier extent, weather station locations, and elevation and hillshade rasters, together with comma-separated value (CSV) tables and matching CSV data dictionaries. These mapped layers provide the geographic framework for interpreting denudation patterns across the Pikes Peak region and for relating sample locations to watershed geometry, bedrock setting, glacial history, and nearby climate stations. The first group of tables reports climate and geospatial context for the study area. These files include station-based temperature and precipitation data used to characterize elevational gradients in mean annual climate and monthly climate seasonality, sample locations, denudation-rate and topographic metrics, fixed frost-cracking model parameters, frost-cracking intensity and precipitation-frequency metrics, and stream-power inversion results. Together, these data provide the basis for evaluating how denudation varies with elevation, climate, and landscape form across sampled catchments on Pikes Peak. The second group of tables reports cosmogenic nuclide and erosion-model results used in the denudation analysis. Included files contain accelerator mass spectrometry (AMS) measurements for in situ-produced cosmogenic beryllium-10 (10Be), including sample identifiers, measured 10Be:9Be ratios, analytical uncertainties, carrier mass, quartz mass, blank corrections, blank-group statistics, and calculated 10Be concentrations and uncertainties. Additional tables summarize stream-power-law inversion results for sampled catchments, including optimized model parameters, predicted erosion rates, residual metrics, channel-pixel counts, and convergence status, as well as regression equations and summary statistics used to evaluate relationships among elevation, climate, frost cracking, precipitation forcing, and denudation rate. The package contains GIS files, comma-separated value files (.csv), Microsoft Excel files (.xlsx), CSV data dictionaries, a file-level metadata table, and a readme text file.

10Be cosmogenic nuclides↗

THE DISTANCE TO M104

M104 (NGC 4594; the Sombrero galaxy) is a nearby, well-studied elliptical galaxy included in scores of surveys focused on understanding the details of galaxy evolution. Despite the importance of observations of M104, a consensus distance has not yet been established. Here, we use newly obtained Hubble Space Telescope optical imaging to measure the distance to M104 based on the tip of the red giant branch (TRGB) method. Our measurement yields the distance to M104 to be 9.55 ± 0.13 ± 0.31 Mpc equivalent to a distance modulus of 29.90 ± 0.03 ± 0.07 mag. Our distance is an improvement over previous results as we use a well-calibrated, stable distance indicator, precision photometry in a optimally selected field of view, and a Bayesian maximum likelihood technique that reduces measurement uncertainties. The most discrepant previous results are due to Tully–Fisher method distances, which are likely inappropriate for M104 given its peculiar morphology and structure. Our results are part of a larger program to measure accurate distances to a sample of well-known spiral galaxies (including M51, M74, and M63) using the TRGB method.

79 ASTRONOMY AND ASTROPHYSICS↗

Assessing the technical aspects of ocean-alkalinity-enhancement approaches

Ocean alkalinity enhancement (OAE) is an emerging strategy that aims to mitigate climate change by increasing the alkalinity of seawater. This approach involves increasing the alkalinity of the ocean to enhance its capacity to absorb and store carbon dioxide (CO 2 ) from the atmosphere. This chapter presents an overview of the technical aspects associated with the full range of OAE methods being pursued and discusses implications for undertaking research on these approaches. Various methods have been developed to implement OAE, including the direct injection of alkaline liquid into the surface ocean; dispersal of alkaline particles from ships, platforms, or pipes; the addition of minerals to coastal environments; and the electrochemical removal of acid from seawater. Each method has its advantages and challenges, such as scalability, cost effectiveness, and potential environmental impacts. The choice of technique may depend on factors such as regional oceanographic conditions, alkalinity source availability, and engineering feasibility. This chapter considers electrochemical methods, the accelerated weathering of limestone, ocean liming, the creation of hydrated carbonates, and the addition of minerals to coastal environments. In each case, the technical aspects of the technologies are considered, and implications for best-practice research are drawn. The environmental and social impacts of OAE will likely depend on the specific technology and the local context in which it is deployed. Therefore, it is essential that the technical feasibility of OAE is undertaken in parallel with, and informed by, wider impact assessments. While OAE shows promise as a potential climate change mitigation strategy, it is essential to acknowledge its limitations and uncertainties. Further research and development are needed to understand the long-term effects, optimize techniques, and address potential unintended consequences. OAE should be viewed as complementary to extensive emission reductions, and its feasibility may be improved if it is operated using energy and supply chains with minimal CO 2 emissions.

54 ENVIRONMENTAL SCIENCES↗

A combined stochastic feedforward and feedback control design methodology with application to autoland design

A combined stochastic feedforward and feedback control design methodology was developed. The objective of the feedforward control law is to track the commanded trajectory, whereas the feedback control law tries to maintain the plant state near the desired trajectory in the presence of disturbances and uncertainties about the plant. The feedforward control law design is formulated as a stochastic optimization problem and is embedded into the stochastic output feedback problem where the plant contains unstable and uncontrollable modes. An algorithm to compute the optimal feedforward is developed. In this approach, the use of error integral feedback, dynamic compensation, control rate command structures are an integral part of the methodology. An incremental implementation is recommended. Results on the eigenvalues of the implemented versus designed control laws are presented. The stochastic feedforward/feedback control methodology is used to design a digital automatic landing system for the ATOPS Research Vehicle, a Boeing 737-100 aircraft. The system control modes include localizer and glideslope capture and track, and flare to touchdown. Results of a detailed nonlinear simulation of the digital control laws, actuator systems, and aircraft aerodynamics are presented.

Halyo, Nesim↗

Robust time-optimal control of uncertain structural dynamic systems

A time-optimal open-loop control problem of flexible spacecraft in the presence of modeling uncertainty has been investigated. The results indicate that the proposed approach significantly reduces the residual structural vibrations caused by modeling uncertainty. The results also indicate the importance of proper jet placement for practical tradeoffs among the maneuvering time, fuel consumption, and performance robustness.

Wie, Bong↗

Confidence of the three-point estimator of frequency drift

It was shown that a three-point second difference estimator is nearly optimal for estimating frequency drift in many common atomic oscillators. A formula for the uncertainty of this estimate as a function of the integration time and of the Allan variance associated with this integration time is derived.

Weiss, Marc A.↗

On the Biases in AIRS Retrieval of Ozone (Work in Progress)

A viewgraph presentation on various biases in AIRS retrieval of ozone is given. The topics include: 1) AIRS captures UTLS ozone events; 2) Simplified AIRS retrieval of ozone; 3) How much is AIRS getting its skill in ozone from regression? biases are similar to ECMWF; 4) How does channel selection and damping affect the retrieval?; 5) Decreasing damping worsens results in upper"trop/lower"strat with current channel selection; 6) If results worse with decreased damping, let s give the retrieval more information; 7) Adding channels at current damping doesn't change anything.; 8) Adding channels and decreasing damping gives mixed results; 9) Let's look at the radiances and their uncertainties ; 10) Systematic biases in radiance uncertainties?; 11) Surface Temperature; 12) Temperature Profile; and 13) x vs Brightness Temperature Error Optimal Estimation Retrieval.

Atmospheric Infrared Sounder (AIRS)↗

Probabilistic Usage of the Multi-Factor Interaction Model

A Multi-Factor Interaction Model (MFIM) is used to predict the insulating foam mass expulsion during the ascending of a space vehicle. The exponents in the MFIM are evaluated by an available approach which consists of least squares and an optimization algorithm. These results were subsequently used to probabilistically evaluate the effects of the uncertainties in each participating factor in the mass expulsion. The probabilistic results show that the surface temperature dominates at high probabilities and the pressure which causes the mass expulsion at low probabil

Chamis, Christos C.↗

Desensitized Optimal Filtering and Sensor Fusion Toolkit

Analytical Mechanics Associates, Inc., has developed a software toolkit that filters and processes navigational data from multiple sensor sources. A key component of the toolkit is a trajectory optimization technique that reduces the sensitivity of Kalman filters with respect to model parameter uncertainties. The sensor fusion toolkit also integrates recent advances in adaptive Kalman and sigma-point filters for non-Gaussian problems with error statistics. This Phase II effort provides new filtering and sensor fusion techniques in a convenient package that can be used as a stand-alone application for ground support and/or onboard use. Its modular architecture enables ready integration with existing tools. A suite of sensor models and noise distribution as well as Monte Carlo analysis capability are included to enable statistical performance evaluations.

Karlgaard, Christopher D.↗

Interval Predictor Models for Robust System Identification

This paper proposes a framework for the identification and uncertainty quantification of plant models according to multivariable data. The only restriction imposed upon such models is for their outputs to depend continuously on their parameters. An Interval Predictor Model (IPM) prescribes the parameters of a computational model as a path-connected set thereby making each predicted output an interval-valued function of its inputs. The formulation proposed seeks the parameter set for which the predicted outputs tightly enclose the data. This set, which is modeled as a semi-algebraic set of low-degree polynomials, enables the characterization of possibly strong parameter dependencies commonly found in practice. This uncertainty characterization makes the resulting plant model amenable to robust control approaches using polynomial optimization. Furthermore, we use non-convex scenario theory to assess the reliability of the resulting IPM. This assessment yields a distribution-free upper bound on the probability that future data will fall outside the predicted intervals.

interval↗

Design of Materials with Alchemite

Machine learning models that establish the relationships between materials processing and properties can enable inverse design of materials through active learning. Alchemite is a commercial software that can perform inverse materials design on sparse data. Here we evaluate Alchemite’s performance on a dataset of shape memory alloys and a dataset of heat exchangers compared to baseline random forest models. Alchemite had higher accuracy when making predictions on sparse data and was more accurate or nearly as accurate as random forests on complete datasets while also quantifying uncertainty. The software was also used to suggest processing steps and design parameters to optimize properties and performance; however, physical validation of the suggested design parameters was beyond the scope of this work. Several useful design insights were gained about the impact of the design parameters on properties and performance including the importance of dopant choice and amount for shape memory alloys and the importance of height and weight on the thermal resistance of heat exchangers.

Machine learning↗

3D Cloud Tomography and Droplet Size Retrieval from Multi-Angle Polarimetric Imaging of Scattered Sunlight from Above

Tomography aims to recover a three-dimensional (3D) density map of a medium or an object. In medical imaging,it is extensively used for diagnostics via X-ray computed tomography (CT). We define and derive a tomographyof cloud droplet distributions via passive remote sensing. We use multi-view polarimetric images to fit a 3Dpolarized radiative transfer (RT) forward model. Our motivation is 3D volumetric probing of vertically-developedconvectively-driven clouds that are ill-served by current methods in operational passive remote sensing. Currenttechniques are indeed based on strictly 1D RT modeling and applied to a single cloudy pixel, where cloud geometrydefaults to that of a plane-parallel slab. Incident unpolarized sunlight, once scattered by cloud droplets, changesits polarization state according to droplet size. Therefore, polarimetric measurements in the rainbow and gloryangular regions can be used to infer the droplet size distribution. This work defines and derives a framework for afull 3D tomography of cloud droplets for both their mass concentration in space and their distribution across arange of sizes. This gridded 3D retrieval of key microphysical properties is made tractable by our novel approachthat involves a restructuring and partial linearization of an open-source polarized 3D RT code to accommodate aspecial two-step iterative optimization technique. Physically-realistic synthetic clouds are used to demonstrate themethodology with rigorous uncertainty quantification, while a real-world cloud imaged by AirMSPI is processedto illustrate the new remote sensing capability

Schechner, Yoav Y.↗

Global Sensitivity Analyses for Test Planning with Black-Box Models for Mars Sample Return

This work describes sensitivity analyses performed on complex black-box models used to support experimental test planning under limited resources in the context of the Mars Sample Return program, which aims at bringing to Earth rock and atmospheric samples from Mars. We develop a systematic workflow that allows the analysts to simultaneously obtain quantitative insights on key drivers of uncertainty, on the direction of impact, and the presence of interactions. We apply novel optimal transport-based global sensitivity measures to tackle the multivariate nature of the output. On the modeling side, we apply multi-fidelity techniques that leverage low-fidelity models to speed up the calculations and make up for the limited amount of high-fidelity samples, while keeping these in the loop for accuracy guarantees. The sensitivity analysis reveals insights useful for the analysts to understand the model's behavior and identify the factors to focus on during testing in order to maximize the value of information extracted from them to ensure mission success when limited resources are available.

Giuseppe Cataldo↗