Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “generalized predictive control”

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 109 records · Page 6

Opening the reheating box in multifield inflation

The robustness of multi-field inflation to the physics of reheating is investigated. In order to carry out this study, reheating is described in detail by means of a formalism which tracks the evolution of scalar fields and perfect fluids in interaction (the inflatons and their decay products). This framework is then used to establish the general equations of motion of the background and perturbative quantities controlling the evolution of the system during reheating. Next, these equations are solved exactly by means of a new numerical code. Moreover, new analytical techniques, allowing us to interpret and approximate these solutions, are developed. As an illustration of a physical prediction that could be affected by the micro-physics of reheating, the amplitude of non-adiabatic perturbations in double inflation is considered. It is found that ignoring the fine-structure of reheating, as usually done in the standard approach, can lead to differences as big as ~ 50%, while our semi-analytic estimates can reduce this error to ~ 10%. Finally, we conclude that, in multi-field inflation, tracking the perturbations through the details of the reheating process is important and, to achieve good precision, requires the use of numerical calculations.

79 ASTRONOMY AND ASTROPHYSICS↗

Artificial Neural Networks for In-Cycle Prediction of Knock Events

Downsized turbocharged engines have been increasingly popular in modern light-duty vehicles due to their fuel efficiency benefits. However, high power density in such engines is achieved thanks to high in-cylinder pressure and temperature conditions that increase knock propensity. Next-cycle control has been studied as a method to reduce the damaging effects of knock by operating the engine in a low knock probability condition. This exploratory study looks at the feasibility of in-cycle knock prediction as a tool for advanced knock control algorithms. A methodology is proposed to 1) choose in-cycle features of the pressure trace that highly correlate with knock events and 2) train artificial neural networks to predict in-cycle knock events before knock onset. The methodology was validated at different operating conditions and different levels of generalization. Precision and recall were used as metrics to evaluate the binary classifier. However, the Fowlkes-Mallows (FM) index was used to compare the result of the clustering algorithm at different operating conditions. The results showed a maximum FM index of 0.7 when the prediction was done at knock onset and a minimum FM index of 0.45 when the prediction was done at spark timing.

42 ENGINEERING↗

GrainNN: A neighbor-aware long short-term memory network for predicting microstructure evolution during polycrystalline grain formation

High fidelity simulations of grain formation in alloys are an indispensable tool for process-to-mechanical-properties characterization. Such simulations, however, can be computationally expensive as they require fine spatial and temporal discretizations. Their cost becomes an obstacle to parametric studies and ensemble runs and ultimately makes downstream tasks like optimal control and uncertainty quantification challenging. To enable such downstream tasks, we introduce GrainNN, an efficient and accurate reduced-order model for epitaxial grain growth in additive manufacturing conditions. GrainNN is a sequence-to-sequence long-short-term-memory (LSTM) deep neural network that evolves the dynamics of manually crafted features. Its innovations are (1) an attention mechanism with grain-microstructure-specific transformer architecture; and (2) an overlapping combination of several clones of the network to generalize to grain configurations that are different from those used for training. This design enables GrainNN to predict grain formation for unseen physical parameters, grain number, domain size and geometry. Furthermore, GrainNN not only reconstructs the quantities of interest but also can be pointwise accurate. In our numerical experiments, we use a polycrystalline phase field method to both generate the training data and assess GrainNN. For multiparametric, ensemble simulations with many grains, GrainNN can be orders of magnitude faster than phase field simulations, while delivering 5%–15% pointwise error. Additionally, this speedup includes the cost of the phase field simulations for generating training data.

36 MATERIALS SCIENCE↗

A modeling pipeline to relate municipal wastewater surveillance and regional public health data

As COVID-19 becomes endemic, public health departments require improved passive indicators, which are independent of voluntary testing data, to estimate the prevalence of COVID-19 in local communities. Quantification of SARS-CoV-2 RNA from wastewater has the potential to be a powerful passive indicator. However, connecting measured SARS-CoV-2 RNA to community prevalence is challenging. We have developed a generalized methodology to improve the predictive power of wastewater measurements and applied it to data collected from treatment plants in the Chicago area. We built and compared a set of multi-linear regression models, which incorporate pepper mild mottle virus (PMMoV) as a population biomarker, Bovine coronavirus (BCoV) as a recovery control, and wastewater system flow rate into a corrected estimate for SARS-CoV-2 RNA concentration. For our data, models with BCoV performed better than those with PMMoV, but all model estimates of prevalence significantly improved correlation compared to doing no correction. We also investigate the utility of RNA measurements in wastewater as a leading indicator of COVID-19 trends. We do this in a rolling manner for corrected wastewater data and for other prevalence indicators, and statistically compare the temporal relationship between new increases in the wastewater data and those in other prevalence indicators. We find that wastewater trends often lead other COVID-19 indicators in predicting new surges.

60 APPLIED LIFE SCIENCES↗

Correlation-aware binning for small-angle neutron scattering via Gaussian-process inference

Binning in small-angle neutron scattering (SANS) is typically performed empirically, with fixed parameters chosen for convenience rather than statistical optimality. Such practices often fail to balance statistical precision and spatial resolution, leading to inconsistencies across instruments and datasets. Here we establish a correlation-aware framework that determines the optimal bin width from first principles by extending the classical Freedman–Diaconis (FD) rule to account for inter-bin correlations with a Gaussian process. In this formulation, the scattering intensity is treated as a smooth stochastic field whose statistical coherence is described by a covariance matrix. Analytical expressions of errors derived from this model yield closed-form criteria that separate the total deviation into contributions from counting noise, aliasing distortion and curvature-dependent correlation effects. Expressed in reduced variables, the resulting dimensionless error surface reveals a continuous transition from the uncorrelated FD regime to the correlation-dominated limit, providing a unified description of noise suppression and resolution control. Because the formulation depends only on the profile characteristics of scattering intensity I(Q), specifically its average intensity and first- and second-order derivatives, it applies generally to any SANS measurement regardless of sample, instrument or geometry. Experimental validation using small- and ultra-small-angle neutron scattering data confirms the predicted scaling behavior, demonstrating that correlation-aware inference systematically reduces mean-squared error and enables information-efficient reproducible data reduction across materials and instruments.

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

Real-time correlation of crystallization and segmental order in conjugated polymers

Modulating the segmental order in the morphology of conjugated polymers is widely recognized as a crucial factor for achieving optimal electronic properties and mechanical deformability. However, it is worth noting that the segmental order is typically associated with the crystallization process, which can result in rigid and brittle long-range ordered crystalline domains. To precisely control the morphology, a comprehensive understanding of how highly anisotropic conjugated polymers form segmentally ordered structures with ongoing crystallization is essential, yet currently elusive. To fill this knowledge gap, we developed a novel approach with a combination of stage-type fast scanning calorimetry and micro-Raman spectroscopy to capture the series of specimens with a continuum in the polymer percent crystallinity and detect the segmental order in real-time. Through the investigation of conjugated polymers with different backbones and side-chain structures, we observed a generally existing phenomenon that the degree of segmental order saturates before the maximum crystallinity is achieved. This disparity allows the conjugated polymers to achieve good charge carrier mobility while retaining good segmental dynamic mobility through the tailored treatment. Moreover, the crystallization temperature to obtain optimal segmental order can be predicted based on T g and T m of conjugated polymers. In conclusion, this in-depth characterization study provides fundamental insights into the evolution of segmental order during crystallization, which can aid in designing and controlling the optoelectronic and mechanical properties of conjugated polymers.

36 MATERIALS SCIENCE↗

Development of Supervisory Control System for Thermal Energy Distribution System

The integrated energy system (IES) refers to the combination of nuclear energy generation with other energy sources to enable the efficiency and reliability of power generations. To create technologically viable and economically competitive systems, supervisory control strategies are critical for optimizing performance and ensuring stability across different energy generation, transportation, and utilization. This work focuses on the control strategies for the thermal energy distribution system (TEDS), which is a cornerstone of the Dynamic Energy Transport and Integration Laboratory at Idaho National Laboratory. TEDS currently relies on operators to coordinate across different components to manage energy storage and ensure efficiency. This work demonstrates the use of model predictive control (MPC) with surrogate models in determining optimal setpoints for major TEDS components. The capability of MPC-based supervisory control system is evaluated by autonomously matching the power outputs from a Dymola-based TEDS with target heat demands.

22 - GENERAL STUDIES OF NUCLEAR REACTORS↗

Observation of Weyl exceptional rings in thermal diffusion

A non-Hermitian Weyl equation indispensably requires a three-dimensional (3D) real/synthetic space, and it is thereby perceived that a Weyl exceptional ring (WER) will not be present in thermal diffusion given its purely dissipative nature. Here, we report a recipe for establishing a 3D parameter space to imitate thermal spinor field. Two orthogonal pairs of spatiotemporally modulated advections are employed to serve as two synthetic parameter dimensions, in addition to the inherent dimension corresponding to heat exchanges. We first predict the existence of WER in our hybrid conduction–advection system and experimentally observe the WER thermal signatures verifying our theoretical prediction. When coupling two WERs of opposite topological charges, the system further exhibits surface-like and bulk topological states, manifested as stationary and continuously changing thermal processes, respectively, with good robustness. Our findings reveal the long-ignored topological nature in thermal diffusion and may empower distinct paradigms for general diffusion and dissipation controls.

42 ENGINEERING↗

Overall chilled water system energy consumption modeling and optimization

The emergence of increasingly affordable variable-speed drive technology has changed the approach used to control chilled water systems equipped with these drives. The purpose of this research was to develop an integrated chilled water modeling technique that can determine the optimal system setpoints and estimate the energy saving potential of chiller system. The chiller system equipped with Variable Frequency Drives (VFDs) on cooling tower fans and condenser water pumps. To accomplish the objective, physical component models of the centrifugal chiller, cooling tower and condenser water pump were established with the goal of incorporating the system’s condenser water flow rate and cooling tower fan speeds as optimization variables. Furthermore, a cooling load prediction algorithm was developed using a multiple non-linear regression model to approximate the building’s cooling load subject to a range of environmental conditions. The inputs and outputs of the individual component models were linked to estimate how adjusting the cooling tower fan and condenser water pump speed would influence the system’s comprehensive performance. Here, the overall system model was then optimized using a generalized reduced gradient optimization algorithm to determine the potential energy savings through speed control with VFDs and to ascertain a control logic strategy for the building automation system to operate the heating and cooling system. A case-study was performed on a single chiller system at a museum and the model was calibrated according to logged data collected over four months. Results showed that for the system analyzed, the energy saving of optimizing the cooling tower fan system was found to be 12–15%, while the energy saving potential of optimizing the condenser water pump with the cooling tower fan was negligible. Additionally, comparing different cooling tower fan control strategies showed that a wet-bulb approach-based cooling tower control strategy was shown to have the highest correlation to the optimized fan speed with an R 2 of 0.924.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

Foundations of Molecular 'Isotomics'

The naturally occurring rare isotopes are versions of common elements, such as hydrogen, carbon and oxygen, that contain a larger than usual number of neutrons in their atomic nuclei and therefore are higher in mass than the common atoms of that element. Isotopes exist for most elements and are found in most natural and synthetic materials, but are uneven in their distribution because chemical and physical processes are isotope-selective (e.g., a chemical reaction may proceed more rapidly for one isotope than for another). For this reason, abundances of isotopes in a material of interest can provide a record, or ‘signature’ of various features of that material’s origin and history. These signatures have been used in the geo, life, chemical and physical sciences in a wide variety of ways over close to 8 decades. However, many such applications struggle to reach unique interpretations of isotopic data because multiple factors combine to control a given sample’s overall isotopic content. That is, the factors controlling isotopic content are too numerous and complex to fully constrain from a simple measurement of a material’s isotope abundances. However, the distribution of isotopes within materials, at molecular scales potentially provides a vastly larger number and diversity of constraints on the chemical and physical processes that comprise a material’s history. The rare isotopes may be concentrated into one atomic position in a molecule relative to another, some proportion of molecules in a sample may contain two or more rare isotopes, and those multiply-isotope-substituted forms of molecules may also have uneven distributions of those isotopes across individual atomic sites. For these reasons, even small, seemingly simple molecules, such as sugars, amino acids or drug compounds, actually exist in a vast number of isotopically unique forms (often millions or more), and each one of those forms is in some sense an independent ‘vote’ on that sample’s history. This project has focused on opening this rich archive of information by enabling the creation of routinely and widely applicable ways of measuring and interpreting isotopic structures of molecules. This work has included the development of core technologies and analytical methods, advancing fundamental understanding of the physical and chemical properties of isotopic versions of molecules, and conducting proof of concept studies of illustrative geochemical, cosmochemical and forensic problems in order to show how these technologies, methods and principles come together to solve problems in new ways. A key to the success of this project was the adaptation of ‘Fourier transform mass spectrometry’ (FTMS) to the task of precisely measuring proportions of the rare, naturally occurring isotopic forms of molecules. FTMS is a highly specialized form of mass spectrometry that traps ions within magnetic or electrostatic cavities and, effectively, ‘listens’ (through registering of subtle electrical signals) to the harmonic signals they make while rapidly orbiting within those cavities. These signals have periods that are a function of their mass and strength (or ‘loudness’) that is proportional to their abundances. Thus, these signals constrain relative amounts of molecules that differ in their mass due to various isotopic substitutions. This technology has been essential to the identification of organic molecules in the life, chemical and environmental sciences for over 4 decades, but generally has lacked the control, stability and precision to meaningfully measure rare isotope forms of molecules. This project’s most fundamental contribution has been to modify FTMS, both in terms of hardware and methods, to enable such measurements. The raw data of molecular isotopic structure is tremendously voluminous and complex, so another important activity of this project has been developing the theoretical and data-science tools needed to interpret the data generated by this new form of isotopic measurement. A particularly challenging part of this task has been predicting molecular isotopic structure, as only through the comparison of measurements with predictions can we make progress on hypothesis driven research questions. We have attacked this this prediction task through a combination of first-principles chemical-physics models of the effects of isotope substitution on molecule properties and data-science models that permit us to generalize that chemical physics to cases that have not yet been studied by detailed chemical physics theory. The proof of concept applications we have pursued over the course of this study include biological reactions of amino acids and other biomolecules, non-biological synthesis of organic molecules in extra-terrestrial settings such as meteorites, petroleum geoscience questions concerning the origin and evolution of natural gas, oil and kerogen compounds, and forensic questions such as the sourcing of chemical weapons. The successes of these applications have laid the groundwork for the next phase of this field’s development, which will include larger scale and more ambitious studies of molecular isotopic structure as a means of diagnosing human diseases, such as cancer, and reconstructing detailed interpretations of the origin and evolution of organic molecules in modern and geological environments.

Cesar, Jaime↗

Autonomous convergence of STM control parameters using Bayesian optimization

Scanning tunneling microscopy (STM) is a widely used tool for atomic imaging of novel materials and their surface energetics. However, the optimization of the imaging conditions is a tedious process due to the extremely sensitive tip–surface interaction, thus limiting the throughput efficiency. In this paper, we deploy a machine learning (ML)-based framework to achieve optimal atomically resolved imaging conditions in real time. The experimental workflow leverages the Bayesian optimization (BO) method to rapidly improve the image quality, defined by the peak intensity in the Fourier space. The outcome of the BO prediction is incorporated into the microscope controls, i.e., the current setpoint and the tip bias, to dynamically improve the STM scan conditions. We present strategies to either selectively explore or exploit across the parameter space. As a result, suitable policies are developed for autonomous convergence of the control parameters. The ML-based framework serves as a general workflow methodology across a wide range of materials.

97 MATHEMATICS AND COMPUTING↗

Developing Drag Models for Non-Spherical Particles through Machine Learning

The overarching goal of this project is to produce comprehensive experimental and numerical datasets for gas-solid flows in well-controlled settings to understand the aerodynamic drag of non-spherical particles in the dense regime. The datasets and the gained knowledge will be utilized to train deep neural networks in TensorFlow to formulate a general drag model for use directly in NETL MFiX-DEM module in order to help to advance the accuracy and prediction fidelity of the computational tools that will be used in designing and optimizing fluidized beds and chemical looping reactors.

42 ENGINEERING↗

Dehydroxylation kinetics of kaolinite and montmorillonite examined using isoconversional methods

The use of calcined clays as supplementary cementitious materials (SCMs) in concrete is a promising strategy towards decarbonizing the cement and concrete industry. This is especially relevant considering the ever-increasing demand for concrete. Comprehensive understanding of the kinetics of calcination is essential towards maximizing the potential reactivity of clay minerals while ensuring energy efficiency. In this study, the kinetics of the dehydroxylation of kaolinite and montmorillonite are investigated under non-isothermal conditions at constant heating rate. Activation energies ( E a ) are determined via Friedman differential and advanced Vyazovkin incremental methods over the isoconversional range; these are devoid of computational approximations, thus allowing kinetic analysis without assuming a specific reaction model. Kinetic equations—in the differential form as well as a combination of differential and integral forms are compared against the experimentally determined reaction models to identify the most probable dehydroxylation mechanism for kaolinite and montmorillonite. A reaction order mechanism is established for dehydroxylation of kaolinite, while montmorillonite is noted to undergo dehydroxylation via a single-step reversible diffusion-controlled process. Kinetic triplet—comprising activation energy, reaction model and pre-exponential factor—is used to predict isothermal calcination conditions, which is further verified using analytical techniques. Heat release rates of clay-portlandite blends from isothermal calorimetry are used within a thermodynamic framework to quantify reactivity of the calcined clays. Here, the study demonstrates a general approach based on isoconversional methods to predict calcination conditions for different clays that can be used in efficient and optimized production of blended cements or SCMs.

36 MATERIALS SCIENCE↗

Dynamic machine learning-based optimization algorithm to improve boiler efficiency

With decreasing computational costs, improvement in algorithms, and the aggregation of large industrial and commercial datasets, machine learning is becoming a ubiquitous tool for process and business innovations. Machine learning is still lacking applications in the field of dynamic optimization for real-time control. This work presents a novel framework for performing constrained dynamic optimization using a recurrent neural network model combined with a metaheuristic optimizer. The framework is designed to augment an existing control system and is purely data-driven, like most industrial Model Predictive Control applications. Several recurrent neural network models are compared as well as several metaheuristic optimizers. Hyperparameters and optimizer parameters are tuned with parameter sweeps, and the resulting values are reported. Further, the best parameters for each optimizer and model combination are demonstrated in closed-loop control of a dynamic simulation, and several recommendations are made for generalizing this framework to other systems. Up to 0.953% improvement is realized over the non-optimized case for a simulated coal-fired boiler. While this is not a large improvement in percentage, the total economic impact is $991,000 per year, and this study builds a foundation for future machine learning with dynamic optimization.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

A general infrastructure for data-driven control design and implementation in tokamaks

A general infrastructure for tokamak controllers based on data-driven neural net models is presented. The paradigm allows for more flexible choices of both the underlying model and the desired controlled variables and targets. The system is implemented and tested on the DIII-D tokamak, enacting simultaneous pressure and temperature control via a finite-set model-predictive controller. Traditional control methods such as proportional–integral–derivative (PID) have proven effective for decoupled control tasks, but scale poorly when trying to achieve more complicated goals such as full state control. This is exactly where model-based controllers succeed.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Controlling Optical Cavities using Machine Learning

The GQuEST experiment endeavors to investigate the unification of Quantum Mechanics and General Relativity within the realm of quantum gravity. This pursuit constitutes a well-established conundrum in the field of physics, characterized by its formidable nature due to the infinitesimal effects manifesting at the Planck scale, which measures a mere 10 35 meters. A remarkable facet of the GQuEST experiment is its capacity to predict observable manifestations of Quantum Gravity on a larger and more accessible scale.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Enhanced physics-constrained deep neural networks for modeling vanadium redox flow battery

Numerical simulation has become indispensable in advancing cost-effective process optimization and control of flow batteries. We propose an enhanced version of the physics-constrained deep neural network (PCDNN) approach to provide high-accuracy voltage predictions in the vanadium redox flow batteries (VRFBs). The purpose of the PCDNN approach is to enforce the physics-based zero-dimensional (0D) VRFB model in a neural network to assure model generalization for various battery operation conditions. However, limited by the simplifications of the 0D model, the PCDNN cannot capture sharp voltage changes in the extreme SOC regions. To improve the accuracy of voltage prediction at extreme ranges, we introduce a second (enhanced) DNN to mitigate the prediction errors carried from the 0D model itself and call the resulting approach enhanced PCDNN (ePCDNN). By comparing with experimental data, we demonstrate that the ePCDNN approach can accurately capture the voltage response throughout the charge–discharge cycle, including the tail region of the voltage discharge curve. The loss function for training the ePCDNN is designed to be flexible by adjusting the weights of the physics-constrained DNN and the enhanced DNN. In conclusion, this allows the ePCDNN framework to be transferable to battery systems with variable physical model fidelity.

25 ENERGY STORAGE↗

Photoredox Catalysis with Spin Magnetic Field Effects: A Scheme for Chiral Resolution

Quantities that break both mirror symmetry and time-reversal symmetry, such as the orbital angular momentum, are known to connect molecular chirality with an applied magnetic field. This concept has led to observations such as magneto-chiral dichroism and chirality-induced spin selectivity (CISS). However, being small, these effects often require additional amplification procedures such as flow chemistry to achieve bulk enantioseparation. In this work, we demonstrate how the magnetic field effect on photogenerated radical pairs, which also breaks time-reversal symmetry, can be harnessed for enantiopurification. Fundamental to this process is the collective decay of the singlet and triplet radical-pair states made possible by an applied magnetic field. Because opposite enantiomers exhibit spin-orbit coupling matrix elements of opposite signs, the singlet and triplet decay channels interfere constructively in one enantiomer. Meanwhile, molecules of the other enantiomer are funnelled into the first enantiomer through excited-state chirality inversion, achieving (dynamic kinetic) chiral resolution. Using an axially chiral binaphthyl derivative and a borane photosensitizer as prototype, we predict an appreciable enantiomeric excess (e.e.) of 90% to be possible at steady state, attained within hundreds of milliseconds when irradiated by a laser. Importantly, our analytical results showcase regimes of perfect enantioselectivity (100% e.e.), accessible by further chemical optimization of the photosensitizer for which general strategies are discussed. Altogether, this work illustrates a so-far untapped but powerful control knob for photoredox catalysis based on spin chemistry principles.

Magnetic properties↗