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 1,117 records · Page 62

Uncertainty quantification in reacting flow.

Chemically reacting flow models generally involve inputs and parameters that are determined from empirical measurements, and therefore exhibit a certain degree of uncertainty. Estimating the propagation of this uncertainty into computational model output predictions is crucial for purposes of reacting flow model validation, model exploration, as well as design optimization. Recent years have seen great developments in probabilistic methods and tools for efficient uncertainty quantification (UQ) in computational models. These tools are grounded in the use of Polynomial Chaos (PC) expansions for representation of random variables. The utility and effectiveness of PC methods have been demonstrated in a range of physical models, including structural mechanics, transport in porous media, fluid dynamics, aeronautics, heat transfer, and chemically reacting flow. While high-dimensionality remains nominally an ongoing challenge, great strides have been made in dealing with moderate dimensionality along with non-linearity and oscillatory dynamics. In this talk, I will give an overview of UQ in chemical systems. I will cover both: (1) the estimation of uncertain input parameters from empirical data, and (2) the forward propagation of parametric uncertainty to model outputs. I will cover the basics of forward PC UQ methods with examples of their use. I will also highlight the need for accurate estimation of the joint probability density over the uncertain parameters, in order to arrive at meaningful estimates of model output uncertainties. Finally, I will discuss recent developments on the inference of this density given partial information from legacy experiments, in the absence of raw data.

99 GENERAL AND MISCELLANEOUS↗

Assessment and Verification of SLS Block 1-B Exploration Upper Stage State and Stage Disposal Performance

One of the SLS Navigation System's key performance requirements is a constraint on the payload system's delta-v allocation to correct for insertion errors due to vehicle state uncertainty at payload separation. The SLS navigation team has developed a Delta-Delta-V analysis approach to assess the effect on trajectory correction maneuver (TCM) design needed to correct for navigation errors. This approach differs from traditional covariance analysis based methods and makes no assumptions with regard to the propagation of the state dynamics. This allows for consideration of non-linearity in the propagation of state uncertainties. The Delta-Delta-V analysis approach re-optimizes perturbed SLS mission trajectories by varying key mission states in accordance with an assumed state error. The state error is developed from detailed vehicle 6-DOF Monte Carlo analysis or generated using covariance analysis. These perturbed trajectories are compared to a nominal trajectory to determine necessary TCM design. To implement this analysis approach, a tool set was developed which combines the functionality of a 3-DOF trajectory optimization tool, Copernicus, and a detailed 6-DOF vehicle simulation tool, Marshall Aerospace Vehicle Representation in C (MAVERIC). In addition to delta-v allocation constraints on SLS navigation performance, SLS mission requirement dictate successful upper stage disposal. Due to engine and propellant constraints, the SLS Exploration Upper Stage (EUS) must dispose into heliocentric space by means of a lunar fly-by maneuver. As with payload delta-v allocation, upper stage disposal maneuvers must place the EUS on a trajectory that maximizes the probability of achieving a heliocentric orbit post Lunar fly-by considering all sources of vehicle state uncertainty prior to the maneuver. To ensure disposal, the SLS navigation team has developed an analysis approach to derive optimal disposal guidance targets. This approach maximizes the state error covariance prior to the maneuver to develop and re-optimize a nominal disposal maneuver (DM) target that, if achieved, would maximize the potential for successful upper stage disposal. For EUS disposal analysis, a set of two tools was developed. The first considers only the nominal pre-disposal maneuver state, vehicle constraints, and an a priori estimate of the state error covariance. In the analysis, the optimal nominal disposal target is determined. This is performed by re-formulating the trajectory optimization to consider constraints on the eigenvectors of the error ellipse applied to the nominal trajectory. A bisection search methodology is implemented in the tool to refine these dispersions resulting in the maximum dispersion feasible for successful disposal via lunar fly-by. Success is defined based on the probability that the vehicle will not impact the lunar surface and will achieve a characteristic energy (C3) relative to the Earth such that it is no longer in the Earth-Moon system. The second tool propagates post-disposal maneuver states to determine the success of disposal for provided trajectory achieved states. This is performed using the optimized nominal target within the 6-DOF vehicle simulation. This paper will discuss the application of the Delta-Delta-V analysis approach for performance evaluation as well as trajectory re-optimization so as to demonstrate the system's capability in meeting performance constraints. Additionally, further discussion of the implementation of assessing disposal analysis will be provided.

Patrick, Sean↗

Distributionally Robust Decision Making Leveraging Conditional Distributions

Distributionally robust optimization (DRO) is a powerful tool for decision making under uncertainty. It is particularly appealing because of its ability to leverage existing data. However, many practical problems call for decision-making with some auxiliary information, and DRO in the context of conditional distributions is not straightforward. We propose a conditional kernel distributionally robust optimization (CKDRO) method that enables robust decision making under conditional distributions through kernel DRO and the conditional mean operator in the reproducing kernel Hilbert space (RKHS). In particular, we consider problems where there is a correlation between the unknown variable y and an auxiliary observable variable x. Given past data of the two variables and a queried auxiliary variable, CKDRO represents the conditional distribution P(y|x) as the conditional mean operator in the RKHS space and quantifies the ambiguity set in the RKHS as well, which depends on the size of the dataset as well as the query point. To justify the use of RKHS, we demonstrate that the ambiguity set defined in RKHS can be viewed as a ball under a metric that is similar to the Wasserstein metric. The DRO is then dualized and solved via a finite dimensional convex program. The proposed CKDRO approach is applied to a generation scheduling problem and shows that the result of CKDRO is superior to common benchmarks in terms of quality and robustness.

Chen, Yuxiao↗

An Integrated Approach To Offshore Wind Energy Assessment: Great Lakes 3D Wind Experiment

This grant supported fundamental research into the characterization of flow parameters of relevance to the wind energy industry focused on offshore and the coastal zone. A major focus of the project was application of the latest generation of remote sensing instrumentation and also integration of measurements and numerical modeling to optimize characterization of time-evolving atmospheric flow parameters in 3-D. Our research developed a new data-constrained Wind Atlas for the Great Lakes, and developed new insights into flow parameters in heterogeneous environments. Four experiments were conducted during the project: At a large operating onshore wind farm in May 2012; At the National Renewable Energy Laboratory National Wind Technology Center (NREL NWTC) during February 2013; At the shoreline of Lake Erie in May 2013; and At the Wind Energy Institute of Canada on Prince Edward Island in May 2015. The experiment we conducted in the coastal zone of Lake Erie indicated very complex flow fields and the frequent presence of upward momentum fluxes and resulting distortion of the wind speed profile at turbine relevant heights due to swells in the Great Lakes. Additionally, our data (and modeling) indicate the frequent presence of low level jets at 600 m height over the Lake and occasions when the wind speed profile across the rotor plane may be impacted by this phenomenon. Experimental data and modeling of the fourth experiment on Prince Edward Island showed that at 10-14 m escarpment adjacent to long-overseas fetch the zone of wind speed decrease before the terrain feature and the increase at (and slightly downwind of) the escarpment is ~3–5% at turbine hub-heights. Additionally, our measurements were used to improve methods to compute the uncertainty in lidar-derived flow properties and to optimize lidar-scanning strategies. For example, on the basis of the experimental data we collected plus those from one of our research partners we advanced a new methodology to estimate a priori the uncertainty in wind speed retrievals from arc scans based on site characteristics such as wind velocity, turbulence intensity and proposed scan geometry. Insights regarding use of remote sensing technologies deriving from project experiments were used to compile a best practice document http://doi.org/10.7298/X4QV3JGF for measuring wind speeds and turbulence offshore through in-situ and remote sensing technologies. A project-specific web-site was developed and is available at: http://www.geo.cornell.edu/eas/PeoplePlaces/Faculty/spryor/DoE_AIATOWEA/index.html

17 WIND ENERGY↗

Optimal Power Flow for Distribution Systems under Uncertain Forecasts

The paper focuses on distribution systems featuring renewable energy sources and energy storage devices, and develops an optimal power flow (OPF) approach to optimize the system operation in spite of forecasting errors. The proposed method builds on a chance-constrained multi-period AC OPF formulation, where probabilistic constraints are utilized to enforce voltage regulation with a prescribed probability. To enable a computationally affordable solution approach, a convex reformulation of the OPF task is obtained by resorting to i) pertinent linear approximations of the power flow equations, and ii) convex approximations of the chance constraints. Particularly, the approximate chance constraints provide conservative bounds that hold for arbitrary distributions of the forecasting errors. An adaptive optimization strategy is then obtained by embedding the proposed OPF task into a model predictive control framework.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Uncertainty quantification in reacting flow.

Chemically reacting flow models generally involve inputs and parameters that are determined from empirical measurements, and therefore exhibit a certain degree of uncertainty. Estimating the propagation of this uncertainty into computational model output predictions is crucial for purposes of reacting flow model validation, model exploration, as well as design optimization. Recent years have seen great developments in probabilistic methods and tools for efficient uncertainty quantification (UQ) in computational models. These tools are grounded in the use of Polynomial Chaos (PC) expansions for representation of random variables. The utility and effectiveness of PC methods have been demonstrated in a range of physical models, including structural mechanics, transport in porous media, fluid dynamics, aeronautics, heat transfer, and chemically reacting flow. While high-dimensionality remains nominally an ongoing challenge, great strides have been made in dealing with moderate dimensionality along with non-linearity and oscillatory dynamics. In this talk, I will give an overview of UQ in chemical systems. I will cover both: (1) the estimation of uncertain input parameters from empirical data, and (2) the forward propagation of parametric uncertainty to model outputs. I will cover the basics of forward PC UQ methods with examples of their use. I will also highlight the need for accurate estimation of the joint probability density over the uncertain parameters, in order to arrive at meaningful estimates of model output uncertainties. Finally, I will discuss recent developments on the inference of this density given partial information from legacy experiments, in the absence of raw data.

99 GENERAL AND MISCELLANEOUS↗

Uncertainty modelling for the control of an active structure

Two approaches to uncertainty modeling of a precision active truss structure for control design are presented. The first approach is based on the H(infinity) optimal control theory, which is used for robust control design with an additive uncertainty model. It is designed to provide vibration suppression to selected modes of the structure at the accelerometer locations. The uncertainty model can be used as an adjustable design parameter which will enable a study of the trade-offs between performance and robust stability. The second approach is based on mu-synthesis. Although the structured singular value can be used to analyze the robustness of any type of uncertainty, there exist difficulties in synthesis if the uncertainty block contains both real and complex variations. Mu-synthesis techniques do not allow for real-parameter variations in the synthesis procedure. The use of a complex variation description enclosing the real variation is suggested.

Blackwood, G. H.↗

Peak Reduction in a Residential Community Through Bayesian Optimization of Transactive Control Signals

Daily residential power consumption in aggregate tends to have a large peak that places stress on the distribution system and induces high operational costs. This work shows that coordinated transactive control of noncritical loads within a residential community or microgrid can help to alleviate peak-time stress on the distribution system by flattening the aggregate load curve. Treating the load forecaster as a high-fidelity, expensive black-box function, a new algorithm utilizing Bayesian optimization (BO) is proposed to achieve the best solution under uncertainty with minimal computing effort. The proposed BO algorithm manipulates the shape of the load based on transactive signals sent to each home. The thermostatically controlled loads (TCLs) act out of self-interest in response to the given price while maintaining comfort, and the optimizer exploits the thermal energy retention of the homes for the benefit of the community. Simulations confirm consistent neighborhood-level peak power reduction and energy cost savings.

Schomer, Ian↗

Beyond pinball loss: Quantile methods for calibrated uncertainty quantification

Amongthemanywaysofquantifying uncertainty in a regression setting, specifying the full quantile function is attractive, as quantiles are amenable to interpretation and evaluation. A model that predicts the true conditional quantiles for each input, at all quantile levels, presents a correct and efficient representation of the underlying uncertainty. To achieve this, many current quantile-based methods focus on optimizing the pinball loss. However, this loss restricts the scope of applicable regression models, limits the ability to target many desirable properties (e.g. calibration, sharpness, centered intervals), and may produce poor conditional quantiles. In this work, we develop new quantile methods that address these shortcomings. In particular, we propose methods that can apply to any class of regression model, select an explicit balance between calibration and sharpness, optimize for calibration of centered intervals, and produce more accurate conditional quantiles. We provide a thorough experimental evaluation of our methods, which includes a high dimensional uncertainty quantification task in nuclear fusion.

97 MATHEMATICS AND COMPUTING↗

Uncertainty quantification of mass models using ensemble Bayesian model averaging

Developments in the description of the masses of atomic nuclei have led to various nuclear mass models that provide predictions for masses across the whole chart of nuclides. These mass models play an important role in understanding the synthesis of heavy elements in the rapid neutron capture ( r ) process. However, it is still a challenging task to estimate the size of uncertainty associated with the predictions of each mass model. In this work, a method called ensemble Bayesian model averaging (EBMA) is introduced to quantify the uncertainty of one-neutron separation energies (S 1 n ) which are directly relevant in the calculations of r -process observables. Here, this Bayesian method provides a natural way to perform model averaging, selection, and uncertainty quantification, by combining the mass models as a mixture of normal distributions whose parameters are optimized against the experimental data, employing the Markov chain Monte Carlo method using the no-u-turn sampler. The EBMA model optimized with all the experimental S 1 n from the AME2003 nuclides are shown to provide reliable uncertainty estimates when tested with the new data in the AME2020.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Aerodynamic Optimization of Rocket Control Surface Geometry Using Cartesian Methods and CAD Geometry

Aerodynamic design is an iterative process involving geometry manipulation and complex computational analysis subject to physical constraints and aerodynamic objectives. A design cycle consists of first establishing the performance of a baseline design, which is usually created with low-fidelity engineering tools, and then progressively optimizing the design to maximize its performance. Optimization techniques have evolved from relying exclusively on designer intuition and insight in traditional trial and error methods, to sophisticated local and global search methods. Recent attempts at automating the search through a large design space with formal optimization methods include both database driven and direct evaluation schemes. Databases are being used in conjunction with surrogate and neural network models as a basis on which to run optimization algorithms. Optimization algorithms are also being driven by the direct evaluation of objectives and constraints using high-fidelity simulations. Surrogate methods use data points obtained from simulations, and possibly gradients evaluated at the data points, to create mathematical approximations of a database. Neural network models work in a similar fashion, using a number of high-fidelity database calculations as training iterations to create a database model. Optimal designs are obtained by coupling an optimization algorithm to the database model. Evaluation of the current best design then gives either a new local optima and/or increases the fidelity of the approximation model for the next iteration. Surrogate methods have also been developed that iterate on the selection of data points to decrease the uncertainty of the approximation model prior to searching for an optimal design. The database approximation models for each of these cases, however, become computationally expensive with increase in dimensionality. Thus the method of using optimization algorithms to search a database model becomes problematic as the number of design variables is increased.

Nelson, Andrea↗

Chance-Constrained AC Optimal Power Flow for Distribution Systems With Renewables

This paper focuses on distribution systems featuring renewable energy sources (RESs) and energy storage systems, and presents an AC optimal power flow (OPF) approach to optimize system-level performance objectives while coping with uncertainty in both RES generation and loads. The proposed method hinges on a chance-constrained AC OPF formulation where probabilistic constraints are utilized to enforce voltage regulation with prescribed probability. A computationally more affordable convex reformulation is developed by resorting to suitable linear approximations of the AC power-flow equations as well as convex approximations of the chance constraints. The approximate chance constraints provide conservative bounds that hold for arbitrary distributions of the forecasting errors. An adaptive strategy is then obtained by embedding the proposed AC OPF task into a model predictive control framework. Finally, a distributed solver is developed to strategically distribute the solution of the optimization problems across utility and customers.

24 POWER TRANSMISSION AND DISTRIBUTION↗

Robust Optimal Adaptive Control Method with Large Adaptive Gain

In the presence of large uncertainties, a control system needs to be able to adapt rapidly to regain performance. Fast adaptation is referred to the implementation of adaptive control with a large adaptive gain to reduce the tracking error rapidly. However, a large adaptive gain can lead to high-frequency oscillations which can adversely affect robustness of an adaptive control law. A new adaptive control modification is presented that can achieve robust adaptation with a large adaptive gain without incurring high-frequency oscillations as with the standard model-reference adaptive control. The modification is based on the minimization of the Y2 norm of the tracking error, which is formulated as an optimal control problem. The optimality condition is used to derive the modification using the gradient method. The optimal control modification results in a stable adaptation and allows a large adaptive gain to be used for better tracking while providing sufficient stability robustness. Simulations were conducted for a damaged generic transport aircraft with both standard adaptive control and the adaptive optimal control modification technique. The results demonstrate the effectiveness of the proposed modification in tracking a reference model while maintaining a sufficient time delay margin.

Nguyen, Nhan T.↗

Uncertainty analysis of axial temperature and Seebeck coefficient measurements

Experimental investigations of solid materials at elevated temperatures rely on the optimized thermal design of the measurement system, as radiation becomes a predominant source of heat loss which can lead to large uncertainty in measured temperature and related physical properties of a test sample. Advancements in surface temperature measurements have reduced thermal losses arising from the cold-finger effect using axially inserted thermocouples and from radiation using shields or other thermal guards. The leading technology for temperature sensing at temperatures up to ~900 °C makes use of these design features for measuring thermopower, yet uncertainty analysis estimation of this technique is not known. This study makes use of finite element modeling to determine spatial temperature distributions to obtain the upper limit of confidence expected for the axially inserted thermocouple approach when a heated radiation shield is incorporated into the design. Using an axially inserted thermocouple to measure the sample surface temperature, the temperature variations across the sample hot and cold surfaces at 900 °C for a temperature drop of 0, 5, and 10 °C are calculated to be as low as 0.02, 0.21, and 0.41 °C, respectively, when a heated radiation shield is employed. Uniform temperature distribution on the thermocouple cross-wire geometry indicates that the axial thermocouple measurement design is indeed effective for suppressing the cold-finger effect. Using a heated radiation shield is found to significantly reduce the temperature gradient across the thermocouples.

30 DIRECT ENERGY CONVERSION↗

An Evaluation of the Economic and Resilience Benefits of a Microgrid in Northampton, Massachusetts

Recent developments and advances in distributed energy resource (DER) technologies make them valuable assets in microgrids. This paper presents an innovative evaluation framework for microgrid assets to capture economic benefits from various grid and behind-the-meter services in grid-connecting mode and resilience benefits in islanding mode. In particular, a linear programming formulation is used to model different services and DER operational constraints to determine the optimal DER dispatch to maximize economic benefits. For the resiliency analysis, a stochastic evaluation procedure is proposed to explicitly quantify the microgrid survivability against a random outage, considering uncertainties associated with photovoltaic (PV) generation, system load, and distributed generator failures. Optimal coordination strategies are developed to minimize unserved energy and improve system survivability, considering different levels of system connectedness. The proposed framework has been applied to evaluate a proposed microgrid in Northampton, Massachusetts that would link the Northampton Department of Public Works, Cooley Dickenson Hospital, and Smith Vocational Area High School. The findings of this analysis indicate that over a 20-year economic life, a 441 kW/441 kWh battery energy storage system, and 386 kW PV solar array can generate $2.5 million in present value benefits, yielding a 1.16 return on investment ratio. Results of this study also show that forming a microgrid generally improves system survivability, but the resilience performance of individual facilities varies depending on power-sharing strategies.

25 ENERGY STORAGE↗

Improving the Quasi‐Biennial Oscillation via a Surrogate‐Accelerated Multi‐Objective Optimization

Accurate simulation of the quasi-biennial oscillation (QBO) is challenging due to uncertainties in representing convectively generated gravity waves. We develop an end-to-end uncertainty quantification workflow that calibrates these gravity wave processes in E3SM for a realistic QBO. Central to our approach is a domain knowledge-informed, compressed representation of high-dimensional spatio-temporal wind fields. By employing a parsimonious statistical model that learns the fundamental frequency from complex observations, we extract interpretable and physically meaningful quantities capturing key attributes. Building on this, we train a probabilistic surrogate model that approximates the fundamental characteristics of the QBO as functions of critical physics parameters governing gravity wave generation. Leveraging the Karhunen–Loève decomposition, our surrogate efficiently represents these characteristics as a set of orthogonal features, capturing cross-correlations among multiple physics quantities evaluated at different pressure levels and enabling rapid surrogate-based inference at a fraction of the computational cost of full-scale simulations. Finally, we analyze the inverse problem using a multi-objective approach. Our study reveals a tension between amplitude and period that constrains the QBO representation, precluding a single optimal solution. To navigate this, we quantify the bi-criteria trade-off and generate a set of Pareto optimal parameter values that balance the conflicting objectives. This integrated workflow improves the fidelity of QBO simulations and offers a versatile template for uncertainty quantification in complex geophysical models.

54 ENVIRONMENTAL SCIENCES↗

Constrained variational optimization of counting-time allocation in sequential scattering measurements: Application to Bonse–Hart USANS

Sequential scattering measurements are often performed under a fixed experimental-time budget, even though the expected count rate varies strongly across the measured coordinate. When the dwell time at each measurement position can be controlled independently, this variation creates a general resource-allocation problem: how should the available time be distributed to minimize the uncertainty of the reconstructed profile? We formulate this problem as a constrained variational optimization for measurements governed by Poisson counting statistics. When each measurement is treated independently, minimizing the averaged squared relative uncertainty yields an inverse-square-root intensity allocation. The formulation is then generalized to include correlations between neighboring measurements and an instrumental resolution operator, leading to an allocation criterion that equalizes the marginal reduction in posterior uncertainty per unit measurement time. Bonse–Hart ultra-small-angle neutron scattering (USANS), in which reciprocal space is sampled sequentially through analyzer-angle stepping, provides an experimentally grounded application. Computational benchmarking shows that the optimized allocation outperforms uniform-time and constant-relative-error strategies, while application to an experimentally measured graphite USANS profile from the Spallation Neutron Source, using Poisson resampling under alternative schedules, demonstrates how counting time should be redistributed toward weak-intensity regions under an identical total duration. The resulting framework applies to sequential scattering and related scanning measurements whenever local dwell times are adjustable and directly determine the measurement uncertainties, and when the relevant correlation and instrumental-response models are available.

Tung, Chi-Huan [ORNL] (ORCID:0000000221972074)↗

Distributionally Robust Variational Quantum Algorithms With Shifted Noise

Given their potential to demonstrate near-term quantum advantage, variational quantum algorithms (VQAs) have been extensively studied. Although numerous techniques have been developed for VQA parameter optimization, it remains a significant challenge. A practical issue is the high sensitivity of quantum noise to environmental changes, and its propensity to shift in real time. This presents a critical problem as an optimized VQA ansatz may not perform effectively under a different noise environment. For the first time, we explore how to optimize VQA parameters to be robust against unknown shifted noise. We model the noise level as a random variable with an unknown probability density function (PDF), and we assume that the PDF may shift within an uncertainty set. This assumption guides us to formulate a distributionally robust optimization problem, with the goal of finding parameters that maintain effectiveness under shifted noise. We utilize a distributionally robust Bayesian optimization solver for our proposed formulation. This provides numerical evidence in both the Quantum Approximate Optimization Algorithm (QAOA) and the Variational Quantum Eigensolver (VQE) with hardware-efficient ansatz, indicating that we can identify parameters that perform more robustly under shifted noise. We regard this work as the first step towards improving the reliability of VQAs influenced by real-time noise.

97 MATHEMATICS AND COMPUTING↗