Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “nonlinear regression”

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 271 records · Page 15
All sourcesNASA NTRSDOE OSTI
DOE OSTI2019

Impact of ethanol blending into gasoline on aromatic compound evaporation and particle emissions from a gasoline direct injection engine

This study examined the interactions of ethanol with aromatic compounds on aromatic species evolution during gasoline evaporation and the consequent impacts on particle emissions from a single-cylinder, gasoline direct injection engine. From a chemical kinetic standpoint, the dilution of the aromatic species in gasoline by ethanol blending should reduce particle emissions because aromatics are the primary species that form particles during gasoline combustion. However, increased evaporative cooling from the ethanol and ethanol's effect on when aromatics in the fuel droplet (or surface film or pool) evaporate can drive an increase in particle emissions. The objective of this study was to quantify these competing effects. The results show that a combination of ethanol's increased evaporative cooling and impact on aromatic compound vapor-liquid equilibrium can extend droplet lifetime and cause aromatics to evaporate later in the droplet evaporation process than would be the case without ethanol. When fuels were burned in a direct-injection single-cylinder engine at high speed (2500 rpm) and relatively high load, ethanol blending caused an increase in particle emissions for fuels containing a low vapor pressure aromatic. At a lower speed (1500 rpm), no statistically significant increase in particle emissions was observed, likely because more time was available for evaporation and mixing to occur. In contrast, particle emissions from a fuel blended with a high vapor pressure aromatic were either insensitive to or reduced by ethanol blending, likewise to engine operating conditions. Current models for predicting particle emissions from gasoline engines, such as particulate matter index (PMI), generally lack variables for non-linear interactions between ethanol and gasoline hydrocarbons. To identify nonlinear interactions between predictor variables that might better explain the observed particle emissions, the least absolute shrinkage and selection operator (LASSO) regularized regression approach was employed. Among the predictor variables evaluated was the Yield Sooting Index (YSI) which is as an improved measure to quantify the aromatic's chemical tendency to generate particle emissions. Following this analysis, two variables were selected based on their frequent appearance. The first was ethanol mole percent multiplied by aromatic mole percent divided by aromatic vapor pressure [(EtOH * Aro%)/Aro VP@443 K]. The second was aromatic mole percent multiplied by aromatic YSI divided by aromatic molecular weight [(Aro% * Aro YSI)/Aro MW]. The results of linear regression including only these two combined explanatory variables achieved an r 2 = 0.959, substantially improved over the PMI model (r 2 = 0.688) and provides insight into how the PMI formalism might be modified to account for non-linear oxygenate-hydrocarbon coupling.

09 BIOMASS FUELS↗
NASA NTRSDate not supplied

Computationally Efficient Frequency Domain Method for Dynamic Control Derivative Estimation with Application to Transonic Truss-Braced Wing

This paper presents a computationally efficient method for dynamic control derivative estimation technique via control surface oscillation numerical experiments. Unsteady RANS CFD simulations in FUN3D are performed to simulate the control surface oscillations with a prescribed truncated square wave containing sufficient frequencies of interest. The truncated square wave oscillation offers the computational efficiency which reduces the computational cost by almost an order of magnitude compared to a sine wave oscillation. The nonlinear effect of large control surface oscillation amplitudes creates a spillover effect whereby the frequency response at the same input frequencies contains not only the linear aerodynamic response but also nonlinear aerodynamic response. A correction procedure is developed to remove the spillover effect from the linear aerodynamic response. A frequency-domain regression is performed to estimate the dynamic control derivatives after the correction. The results generally agree with the previous results obtained from the sine wave oscillation.

Stability and Control↗
DOE OSTI2025

Bayesian learning with Gaussian processes for low-dimensional representations of time-dependent nonlinear systems

This work presents a data-driven method for learning low-dimensional time-dependent physics-based surrogate models whose predictions are endowed with uncertainty estimates. We use the operator inference approach to model reduction that poses the problem of learning low-dimensional model terms as a regression of state space data and corresponding time derivatives by minimizing the residual of reduced system equations. Standard operator inference models perform well with accurate training data that are dense in time, but producing stable and accurate models when the state data are noisy and/or sparse in time remains a challenge. Another challenge is the lack of uncertainty estimation for the predictions from the operator inference models. Our approach addresses these challenges by incorporating Gaussian process surrogates into the operator inference framework to (1) probabilistically describe uncertainties in the state predictions and (2) procure analytical time derivative estimates with quantified uncertainties. The formulation leads to a generalized least-squares regression and, ultimately, reduced-order models that are described probabilistically with a closed-form expression for the posterior distribution of the operators. The resulting probabilistic surrogate model propagates uncertainties from the observed state data to reduced-order predictions. Furthermore, we demonstrate the method is effective for constructing low-dimensional models of two nonlinear partial differential equations representing a compressible flow and a nonlinear diffusion–reaction process, as well as for estimating the parameters of a low-dimensional system of nonlinear ordinary differential equations representing compartmental models in epidemiology.

Data-driven model reduction↗
DOE OSTI2025

Nonlinear behavior of urban flood peaks in the U.S. Mid-Atlantic region

Urbanization, i.e., increasing urban development areas in a watershed, is well known as a major cause of increasing flood magnitudes. This study analyzes the observed flood peaks at 262 watersheds in the U.S. Mid-Atlantic region with varying levels of urban development and free from reservoir impacts. Our analysis reveals an interesting, V-shaped nonlinear behavior: flood peaks first decrease and then increase with increasing percentage of urban development area at the watershed scale (PDAW), with the shift occurring at a PDAW threshold of around 10%. Regression analyses suggest that the V-shaped pattern primarily results from complex interactions among climate conditions (e.g., storm-event rainfall) and landscape properties (e.g., elevation, distance to the coast). A neural network model was then developed to capture such interactions, satisfactorily reproducing the V-shaped pattern with an R-squared value of 0.58, RMSE of 6.72 mm/day, and NSE of 0.55. These findings highlight the need to account for nonlinear dynamics in flood prediction and management in the coastal environment.

flood peaks↗
DOE OSTI2022

The optimal beam-loading in two-bunch nonlinear plasma wakefield accelerators

Abstract Due to the highly nonlinear nature of the beam-loading, it is currently not possible to analytically determine the beam parameters needed in a two-bunch plasma wakefield accelerator for maintaining a low energy spread. Therefore in this paper, by using the Broyden–Fletcher–Goldfarb–Shanno algorithm for the parameter scanning with the code QuickPIC and the polynomial regression together with k -fold cross-validation method, we obtain two fitting formulas for calculating the parameters of tri-Gaussian electron beams when minimizing the energy spread based on the beam-loading effect in a nonlinear plasma wakefield accelerator. One formula allows the optimization of the normalized charge per unit length of a trailing beam to achieve the minimal energy spread, i.e. the optimal beam-loading. The other one directly gives the transformer ratio when the trailing beam achieves the optimal beam-loading. A simple scaling law for charges of drive beams and trailing beams is obtained from the fitting formula, which indicates that the optimal beam-loading is always achieved for a given charge ratio of the two beams when the length and separation of two beams and the plasma density are fixed. The formulas can also help obtain the optimal plasma densities for the maximum accelerated charge and the maximum acceleration efficiency under the optimal beam-loading respectively. These two fitting formulas will significantly enhance the efficiency for designing and optimizing a two-bunch plasma wakefield acceleration stage.

Physics↗
DOE OSTI2020

Adaptive Activation Functions Accelerate Convergence in Deep and Physics-informed Neural Networks

We employ adaptive activation functions for regression in deep and physics-informed neural networks (PINNs) to approximate smooth and discontinuous functions as well as solutions of linear and nonlinear partial differential equations. In particular, we solve the nonlinear Klein-Gordon equation, which has smooth solutions, the nonlinear Burgers equation, which can admit high gradient solutions, and the Helmholtz equation. We introduce a scalable hyper-parameter in the activation function, which can be optimized to achieve best performance of the network as it changes dynamically the topology of the loss function involved in the optimization process. The adaptive activation function has better learning capabilities than the traditional one (fixed activation) as it improves greatly the convergence rate, especially at early training, as well as the solution accuracy. To better understand the learning process, we plot the neural network solution in the frequency domain to examine how the network captures successively different frequency bands present in the solution. We consider both forward problems, where the approximate solutions are obtained, as well as inverse problems, where parameters involved in the governing equation are identified. Our simulation results show that the proposed method is a very simple and effective approach to increase the efficiency, robustness and accuracy of the neural network approximation of nonlinear functions as well as solutions of partial differential equations, especially for forward problems. We theoretically prove that in the proposed method, gradient descent algorithms are not attracted to suboptimal critical points or local minima.

machine leaning, Bad minima, Inverse problems, Phy↗
DOE OSTI2021

Bilinear dynamic mode decomposition for quantum control

Abstract Data-driven methods for establishing quantum optimal control (QOC) using time-dependent control pulses tailored to specific quantum dynamical systems and desired control objectives are critical for many emerging quantum technologies. We develop a data-driven regression procedure, bilinear dynamic mode decomposition (biDMD), that leverages time-series measurements to establish quantum system identification for QOC. The biDMD optimization framework is a physics-informed regression that makes use of the known underlying Hamiltonian structure. Further, the biDMD can be modified to model both fast and slow sampling of control signals, the latter by way of stroboscopic sampling strategies. The biDMD method provides a flexible, interpretable, and adaptive regression framework for real-time, online implementation in quantum systems. Further, the method has strong theoretical connections to Koopman theory, which approximates nonlinear dynamics with linear operators. In comparison with many machine learning paradigms minimal data is needed to construct a biDMD model, and the model is easily updated as new data is collected. We demonstrate the efficacy and performance of the approach on a number of representative quantum systems, showing that it also matches experimental results.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2023

gLaSDI: Parametric physics-informed greedy latent space dynamics identification

A parametric adaptive physics-informed greedy Latent Space Dynamics Identification (gLaSDI) method is proposed for accurate, efficient, and robust data-driven reduced-order modeling of high-dimensional nonlinear dynamical systems. In the proposed gLaSDI framework, an autoencoder discovers intrinsic nonlinear latent representations of high-dimensional data, while dynamics identification (DI) models capture local latent-space dynamics. Here, an interactive training algorithm is adopted for the autoencoder and local DI models, which enables identification of simple latent-space dynamics and enhances accuracy and efficiency of data-driven reduced-order modeling. To maximize and accelerate the exploration of the parameter space for the optimal model performance, an adaptive greedy sampling algorithm integrated with a physics-informed residual-based error indicator and random-subset evaluation is introduced to search for the optimal training samples on the fly. Further, to exploit local latent-space dynamics captured by the local DI models for an improved modeling accuracy with a minimum number of local DI models in the parameter space, a -nearest neighbor convex interpolation scheme is employed. The effectiveness of the proposed framework is demonstrated by modeling various nonlinear dynamical problems, including Burgers equations, nonlinear heat conduction, and radial advection. The proposed adaptive greedy sampling outperforms the conventional predefined uniform sampling in terms of accuracy. Compared with the high-fidelity models, gLaSDI achieves 17 to 2,658× speed-up with 1 to 5% relative errors.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2017

Prediction of thyroid C-cell carcinogenicity after chronic administration of GLP1-R agonists in rodents

Increased incidence of C-cell carcinogenicity has been observed for glucagon-like-protein-1 receptor (GLP-1r) agonists in rodents. It is suggested that the duration of exposure is an indicator of carcinogenic potential in rodents of the different products on the market. Furthermore, the role of GLP-1-related mechanisms in the induction of C-cell carcinogenicity has gained increased attention by regulatory agencies. This study proposes an integrative pharmacokinetic/pharmacodynamic (PKPD) framework to identify explanatory factors and characterize differences in carcinogenic potential of the GLP-1r agonist products. PK models for four products (exenatide QW (once weekly), exenatide BID (twice daily), liraglutide and lixisenatide) were developed using nonlinear mixed effects modelling. Predicted exposure was subsequently linked to GLP-1r stimulation using in vitro GLP-1r potency data. A logistic regression model was then applied to exenatide QW and liraglutide data to assess the relationship between GLP-1r stimulation and thyroid C-cell hyperplasia incidence as pre-neoplastic predictor of a carcinogenic response. The model showed a significant association between predicted GLP-1r stimulation and C-cell hyperplasia after 2 years of treatment. The predictive performance of the model was evaluated using lixisenatide, for which hyperplasia data were accurately described during the validation step. The use of a model-based approach provided insight into the relationship between C-cell hyperplasia and GLP-1r stimulation for all four products, which is not possible with traditional data analysis methods. It can be concluded that both pharmacokinetics (exposure) and pharmacodynamics (potency for GLP-1r) factors determine C-cell hyperplasia incidence in rodents. Our work highlights the pharmacological basis for GLP-1r agonist-induced C-cell carcinogenicity. The concept is promising for application to other drug classes. - Highlights: • An integrative PKPD model is applied to study GLP-1r agonist carcinogenicity. • C-cell carcinogenicity is impacted by both pharmacokinetics and pharmacodynamics. • The relation of GLP-1r stimulation and C-cell hyperplasia appears drug-independent. • Understanding carcinogenic risk needs a pharmacological basis.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2022

Time-series forecasting using manifold learning, radial basis function interpolation, and geometric harmonics

We address a three-tier numerical framework based on nonlinear manifold learning for the forecasting of high-dimensional time series, relaxing the “curse of dimensionality” related to the training phase of surrogate/machine learning models. At the first step, we embed the high-dimensional time series into a reduced low-dimensional space using nonlinear manifold learning (local linear embedding and parsimonious diffusion maps). Then, we construct reduced-order surrogate models on the manifold (here, for our illustrations, we used multivariate autoregressive and Gaussian process regression models) to forecast the embedded dynamics. Finally, we solve the pre-image problem, thus lifting the embedded time series back to the original high-dimensional space using radial basis function interpolation and geometric harmonics. The proposed numerical data-driven scheme can also be applied as a reduced-order model procedure for the numerical solution/propagation of the (transient) dynamics of partial differential equations (PDEs). In conclusion, we assess the performance of the proposed scheme via three different families of problems: (a) the forecasting of synthetic time series generated by three simplistic linear and weakly nonlinear stochastic models resembling electroencephalography signals, (b) the prediction/propagation of the solution profiles of a linear parabolic PDE and the Brusselator model (a set of two nonlinear parabolic PDEs), and (c) the forecasting of a real-world data set containing daily time series of ten key foreign exchange rates spanning the time period 3 September 2001–29 October 2020.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
NASA NTRS2020

Global Aerodynamic Modeling Using Automated Local Model Networks in Real Time

A novel method is presented for automated real-time global aerodynamic modeling using local model networks, known as Smoothed Partitioning with Localized Trees in Real Time (SPLITR), as part of NASA’s Learn-to-Fly technology development initiative. The global nonlinear aerodynamics are partitioned into several local regions known as cells, with the dimension, location, and timing of each partition automatically selected based on a residual characterization procedure, under the constraints of real-time operation. Regression trees represent the successive partitioning of the global flight envelope and describe the evolution of the cell structure. Recursive equation-error least-squares parameter estimation in the time domain is used to estimate a model that represents the local aerodynamics in each region, so that it can be updated independently with non-contiguous data in the range of each cell over time. A weighted superposition of these piecewise local models across the flight envelope forms a global nonlinear model that also accurately captures the local aerodynamics. The SPLITR approach is demonstrated using both simulation and flight data, and the results are analyzed in terms of model predictive capabilities as well as interpretability. The results show that SPLITR can be used to automatically partition complex nonlinear aerodynamic behavior, produce an accurate model, and provide valuable physical insight into the local and global aerodynamics.

Weinstein, Rose↗
DOE OSTI2018

The concentration-response between long-term PM2.5 exposure and mortality; A meta-regression approach

Highlights: • PM2.5-mortality effect is significant below 10 μg/m3 and above 20 μg/m3. • Better exposure estimates result in higher effect size estimates. • More control for SES results in higher effect size estimates. • A nonlinear dose-response is seen. • Effects at higher exposures are larger than predicted by the IER/GBD. Long-term exposure to ambient fine particulate matter (≤ 2.5 μg/m{sup 3} in aerodynamic diameter; PM{sub 2.5}) is significantly associated with increased risk of premature mortality. Our goal was to provide an updated meta-analysis of all-cause and cause-specific mortality associated with exposure to PM{sub 2.5} and to better estimate the risk of death as a function of air pollution levels.

54 ENVIRONMENTAL SCIENCES↗
NASA NTRS1990

Error analysis for reducing noisy wide-gap concentric cylinder rheometric data for nonlinear fluids - Theory and applications

This work discusses the propagation of errors for the recovery of the shear rate from wide-gap concentric cylinder viscometric measurements of non-Newtonian fluids. A least-square regression of stress on angular velocity data to a system of arbitrary functions is used to propagate the errors for the series solution to the viscometric flow developed by Krieger and Elrod (1953) and Pawlowski (1953) ('power-law' approximation) and for the first term of the series developed by Krieger (1968). A numerical experiment shows that, for measurements affected by significant errors, the first term of the Krieger-Elrod-Pawlowski series ('infinite radius' approximation) and the power-law approximation may recover the shear rate with equal accuracy as the full Krieger-Elrod-Pawlowski solution. An experiment on a clay slurry indicates that the clay has a larger yield stress at rest than during shearing, and that, for the range of shear rates investigated, a four-parameter constitutive equation approximates reasonably well its rheology. The error analysis presented is useful for studying the rheology of fluids such as particle suspensions, slurries, foams, and magma.

Borgia, Andrea↗
NASA NTRS1974

Estimation of longitudinal aerodynamic coefficients and comparison with wind-tunnel values

Some recent experience at Ames Research Center in the estimation of aerodynamic coefficients for the Lear-Jet and the Augmentor Wing Jet STOL Research Aircraft is reviewed. The coefficients estimated from flight data are compared with values based on large-scale wind-tunnel tests. The results obtained by the regression and quasilinearization identification techniques are also compared. The regression method generally provides the lower standard deviation in the coefficient estimates and provides the better fit to the wind-tunnel values. The addition of nonlinear terms in the aerodynamic equations decreases the difference between the estimated and measured time histories but also increases the standard deviation in the estimated coefficient values.

Wingrove, R. C.↗
DOE OSTI2021

Manifold learning for coarse-graining atomistic simulations: Application to amorphous solids

In this work, we introduce a generalized machine learning framework to probabilistically parameterize upper-scale models in the form of nonlinear PDEs consistent with a continuum theory, based on coarse-grained atomistic simulation data of mechanical deformation and flow processes. The proposed framework utilizes a hypothesized coarse-graining methodology with manifold learning and surrogate-based optimization techniques. Coarse-grained high-dimensional data describing quantities of interest of the multiscale models are projected onto a nonlinear manifold whose geometric and topological structure is exploited for measuring behavioral discrepancies in the form of manifold distances. A surrogate model is constructed using Gaussian process regression to identify a mapping between stochastic parameters and distances. Derivative-free optimization is employed to adaptively identify a unique set of parameters of the upper-scale model capable of rapidly reproducing the system's behavior while maintaining consistency with coarse-grained atomic-level simulations. The proposed method is applied to learn the parameters of the shear transformation zone (STZ) theory of plasticity that describes plastic deformation in amorphous solids as well as coarse-graining parameters needed to translate between atomistic and continuum representations. We show that the methodology is able to successfully link coarse-grained microscale simulations to macroscale observables and achieve a high-level of parity between the models across scales.

36 MATERIALS SCIENCE↗
DOE OSTI2022

Reduced order modeling for flow and transport problems with Barlow Twins self-supervised learning

Abstract We propose a unified data-driven reduced order model (ROM) that bridges the performance gap between linear and nonlinear manifold approaches. Deep learning ROM (DL-ROM) using deep-convolutional autoencoders (DC–AE) has been shown to capture nonlinear solution manifolds but fails to perform adequately when linear subspace approaches such as proper orthogonal decomposition (POD) would be optimal. Besides, most DL-ROM models rely on convolutional layers, which might limit its application to only a structured mesh. The proposed framework in this study relies on the combination of an autoencoder (AE) and Barlow Twins (BT) self-supervised learning, where BT maximizes the information content of the embedding with the latent space through a joint embedding architecture. Through a series of benchmark problems of natural convection in porous media, BT–AE performs better than the previous DL-ROM framework by providing comparable results to POD-based approaches for problems where the solution lies within a linear subspace as well as DL-ROM autoencoder-based techniques where the solution lies on a nonlinear manifold; consequently, bridges the gap between linear and nonlinear reduced manifolds. We illustrate that a proficient construction of the latent space is key to achieving these results, enabling us to map these latent spaces using regression models. The proposed framework achieves a relative error of 2% on average and 12% in the worst-case scenario (i.e., the training data is small, but the parameter space is large.). We also show that our framework provides a speed-up of $$7 \times 10^{6}$$ 7 × 10 6 times, in the best case, and $$7 \times 10^{3}$$ 7 × 10 3 times on average compared to a finite element solver. Furthermore, this BT–AE framework can operate on unstructured meshes, which provides flexibility in its application to standard numerical solvers, on-site measurements, experimental data, or a combination of these sources.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Development of Modified Perturbation Solutions to the One-Phase Stefan Problems With a Convective Boundary

The classical Stefan problem is used to track the moving solid-liquid interface during the freezing process. Perturbation theory has often been applied to find an approximate analytical solution due to the nonlinearity of the moving interface. However, the Stefan number (i.e., the sensible over latent heat) must be small and usually less than 0.01 to assume the perturbation expansion, which in turn limits the thermal engineering applications. In this study, a modified perturbation solution is developed by adding a correction term after the leading-order solution to be valid for a much wider range of Stefan numbers (i.e., 0.01 less than or equal to Ste less than or equal to 1). Specifically, a one-phase Stefan problem is first formulated subjected to a convective boundary in the Cartesian, cylindrical, and spherical coordinate systems for inward solidification. The leading-order solution is calculated based on the regular perturbation theory, while the correction term is obtained using the Monte-Carlo method and a multi-variant regression. Results show that the correction term has a linear relationship with the Stefan number and is not significantly influenced by the Biot number. The proposed modified perturbation solution can accurately and rapidly predict the nonlinear moving interface motion for the freezing process.

DIRECT ENERGY CONVERSION,MATHEMATICS AND COMPUTING↗
DOE OSTI2021

Accelerating phase-field-based microstructure evolution predictions via surrogate models trained by machine learning methods

Abstract The phase-field method is a powerful and versatile computational approach for modeling the evolution of microstructures and associated properties for a wide variety of physical, chemical, and biological systems. However, existing high-fidelity phase-field models are inherently computationally expensive, requiring high-performance computing resources and sophisticated numerical integration schemes to achieve a useful degree of accuracy. In this paper, we present a computationally inexpensive, accurate, data-driven surrogate model that directly learns the microstructural evolution of targeted systems by combining phase-field and history-dependent machine-learning techniques. We integrate a statistically representative, low-dimensional description of the microstructure, obtained directly from phase-field simulations, with either a time-series multivariate adaptive regression splines autoregressive algorithm or a long short-term memory neural network. The neural-network-trained surrogate model shows the best performance and accurately predicts the nonlinear microstructure evolution of a two-phase mixture during spinodal decomposition in seconds, without the need for “on-the-fly” solutions of the phase-field equations of motion. We also show that the predictions from our machine-learned surrogate model can be fed directly as an input into a classical high-fidelity phase-field model in order to accelerate the high-fidelity phase-field simulations by leaping in time. Such machine-learned phase-field framework opens a promising path forward to use accelerated phase-field simulations for discovering, understanding, and predicting processing–microstructure–performance relationships.

36 MATERIALS SCIENCE↗