Search NASA⌕ Search

SEARCH · Search NASA

Results for “Network parameter error”

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 433 records · Page 24

The effect of clock, media, and station location errors on Doppler measurement accuracy

Doppler tracking by the Deep Space Network (DSN) is the primary radio metric data type used by navigation to determine the orbit of a spacecraft. The accuracy normally attributed to orbits determined exclusively with Doppler data is about 0.5 microradians in geocentric angle. Recently, the Doppler measurement system has evolved to a high degree of precision primarily because of tracking at X-band frequencies (7.2 to 8.5 GHz). However, the orbit determination system has not been able to fully utilize this improved measurement accuracy because of calibration errors associated with transmission media, the location of tracking stations on the Earth's surface, the orientation of the Earth as an observing platform, and timekeeping. With the introduction of Global Positioning System (GPS) data, it may be possible to remove a significant error associated with the troposphere. In this article, the effect of various calibration errors associated with transmission media, Earth platform parameters, and clocks are examined. With the introduction of GPS calibrations, it is predicted that a Doppler tracking accuracy of 0.05 microradians is achievable.

Miller, J. K.↗

Payload Performance of Third Generation TDRS and Future Services

NASA has accepted two of the 3rd generation Tracking and Data Relay Satellites, TDRS K, L, and M, designed and built by Boeing Defense, Space & Security (DSS). TDRS K, L, and M provide S-band Multiple Access (MA) service and S-band, Ku-band and Ka-band Single Access (SA) services to near Earth orbiting satellites. The TDRS KLM satellites offer improved services relative to the 1st generation TDRS spacecraft, such as: an enhanced MA service featuring increased EIRPs and G/T; and Ka-band SA capability which provides a 225 and 650 MHz return service (customer-to-TDRS direction) bandwidth and a 50 MHz forward service (TDRS-to-customer direction) bandwidth. MA services are provided through a 15 element forward phased array that forms up to two beams with onboard active beamforming and a 32 element return phased array supported by ground-based beamforming. SA services are provided through two 4.6m tri-band reflector antennas which support program track pointing and autotrack pointing. Prior to NASAs acceptance of the satellites, payload on-orbit testing was performed on each satellite to determine on-orbit compliance with design requirements. Performance parameters evaluated include: EIRP, G/T, antenna gain patterns, SA antenna autotrack performance, and radiometric tracking performance. On-orbit antenna calibration and pointing optimization was also performed on the MA and SA antennas including 24 hour duration tests to characterize and calibrate out diurnal effects. Bit-Error-Rate (BER) tests were performed to evaluate the end-to-end link BER performance of service through a TDRS K and L spacecraft. The TDRS M is planned to be launched in August 2017. This paper summarizes the results of the TDRS KL communications payload on-orbit performance verification and end-to-end service characterization and compares the results with the performance of the 2nd generation TDRS J. The paper also provides a high-level overview of an optical communications application that will augment the data rates supported by the Space Network.

RF↗

Payload Performance of TDRS KL and Future Services

NASA has accepted two of the 3nd generation Tracking and Data Relay Satellites, TDRS K, L, and M, designed and built by Boeing Defense, Space Security (DSS). TDRS K, L, and M provide S-band Multiple Access (MA) service and S-band, Ku-band and Ka-band Single Access (SA) services to near Earth orbiting satellites. The TDRS KLM satellites offer improved services relative to the 1st generation TDRS spacecraft, such as: an enhanced MA service featuring increased EIRPs and GT; and Ka-band SA capability which provides a 225 and 650 MHz return service (customer-to-TDRS direction) bandwidth and a 50 MHz forward service (TDRS-to-customer direction) bandwidth. MA services are provided through a 15 element forward phased array that forms up to two beams with onboard active beamforming and a 32 element return phased array supported by ground-based beamforming. SA services are provided through two 4.6m tri-band reflector antennas which support program track pointing and autotrack pointing. Prior to NASAs acceptance of the satellites, payload on-orbit testing was performed on each satellite to determine on-orbit compliance with design requirements. Performance parameters evaluated include: EIRP, GT, antenna gain patterns, SA antenna autotrack performance, and radiometric tracking performance. On-orbit antenna calibration and pointing optimization was also performed on the MA and SA antennas including 24 hour duration tests to characterize and calibrate out diurnal effects. Bit-Error-Rate (BER) tests were performed to evaluate the end-to-end link BER performance of service through a TDRS K and L spacecraft. The TDRS M is planned to be launched in August 2017. This paper summarizes the results of the TDRS KL communications payload on-orbit performance verification and end-to-end service characterization and compares the results with the performance of the 2nd generation TDRS J. The paper also provides a high-level overview of an optical communications application that will augment the data rates supported by the Space Network.

Laser↗

Fouling modeling and prediction approach for heat exchangers using deep learning

In this article, we develop a generalized and scalable statistical model for accurate prediction of fouling resistance using commonly measured parameters of industrial heat exchangers. This prediction model is based on deep learning where a scalable algorithmic architecture learns non-linear functional relationships between a set of target and predictor variables from large number of training samples. Here, the efficacy of this modeling approach is demonstrated for predicting fouling in an analytically modeled cross-flow heat exchanger, designed for waste heat recovery from flue-gas using room temperature water. The performance results of the trained models demonstrate that the mean absolute prediction errors are under 10 –4 KW –1 for flue-gas side, water side and overall fouling resistances. The coefficients of determination (R 2 ), which characterize the goodness of fit between the predictions and observed data, are over 99%. Even under varying levels of measurement noise in the inputs, we demonstrate that predictions over an ensemble of multiple neural networks achieves better accuracy and robustness to noise. We find that the proposed deep-learning fouling prediction framework learns to follow heat exchanger flow and heat transfer physics, which we confirm using locally interpretable model agnostic explanations around randomly selected operating points. Overall, we provide a robust algorithmic framework for fouling prediction that can be generalized and scaled to various types of industrial heat exchangers.

42 ENGINEERING↗

Investigation of 3D printed lightweight hybrid composites via theoretical modeling and machine learning

Hybrid composites combine two or more different fillers to achieve multifunctional or advanced material properties, such as lightweight and enhanced mechanical properties. The properties of the composites significantly depend on their microstructures, which can be tailored via advanced 3D printing processes. Understanding the process-structure-property relationships is critical to enable the design and engineering of novel hybrid composites for applications in aerospace, automotive, and protective coatings. Here, for this work, we develop 3D printable and lightweight hybrid composites and leverage the conventional design of experiments, a theoretical hybrid model, and an image-driven machine learning (ML) method to investigate their mechanical behaviors. The hybrid composites are formulated with elastomer matrix, microfillers, and thin-shell particles, enabling a significant degree of design freedom of microstructures with densities and mechanical properties varying up to 70% and 91%, respectively. Our statistical analysis indicates that the 3D printing path direction and the microfibers fraction are dominating process parameters with contribution percentages of 45.3% and 57.7% on the specific stiffness and strength, respectively. A hybrid mechanics model is developed based on a simple Weibull distribution function and classical single-filler models to effectively capture the variations in mechanical properties, however, it overestimates the values due to its statistical constraints and idealization of experimental uncertainty. The image-driven ML model leverages the microscale images directly without losing the structural details, shows more accurate predictions with experimental data, and has 48.6% lower root mean square error than the theoretical model.

3D printing↗

Scaling methods for accelerating kinetic Monte Carlo simulations of chemical reaction networks

Various kinetic Monte Carlo algorithms become inefficient when some of the population sizes in a system are large, which gives rise to a large number of reaction events per unit time. In this work, we present a new acceleration algorithm based on adaptive and heterogeneous scaling of reaction rates and stoichiometric coefficients. The algorithm is conceptually related to the commonly used idea of accelerating a stochastic simulation by considering a subvolume λΩ (0 < λ < 1) within a system of interest, which reduces the number of reaction events per unit time occurring in a simulation by a factor 1/λ at the cost of greater error in unbiased estimates of first moments and biased overestimates of second moments. Our new approach offers two unique benefits. First, scaling is adaptive and heterogeneous, which eliminates the pitfall of overaggressive scaling. Second, there is no need for an a priori classification of populations as discrete or continuous (as in a hybrid method), which is problematic when discreteness of a chemical species changes during a simulation. The method requires specification of only a single algorithmic parameter, N c , a global critical population size above which populations are effectively scaled down to increase simulation efficiency. The method, which we term partial scaling, is implemented in the open-source BioNetGen software package. We demonstrate that partial scaling can significantly accelerate simulations without significant loss of accuracy for several published models of biological systems. Finally, these models characterize activation of the mitogen-activated protein kinase ERK, prion protein aggregation, and T-cell receptor signaling.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Physics-informed neural network with transfer learning (TL-PINN) based on domain similarity measure for prediction of nuclear reactor transients

Nuclear reactor safety and efficiency can be enhanced through the development of accurate and fast methods for prediction of reactor transient (RT) states. Physics informed neural networks (PINNs) leverage deep learning methods to provide an alternative approach to RT modeling. Applications of PINNs in monitoring of RTs for operator support requires near real-time model performance. However, as with all machine learning models, development of a PINN involves time-consuming model training. Here, we show that a transfer learning (TL-PINN) approach achieves significant performance gain, as measured by reduction of the number of iterations for model training. Using point kinetic equations (PKEs) model with six neutron precursor groups, constructed with experimental parameters of the Purdue University Reactor One (PUR-1) research reactor, we generated different RTs with experimentally relevant range of variables. The RTs were characterized using Hausdorff and Fréchet distance. We have demonstrated that pre-training TL-PINN on one RT results in up to two orders of magnitude acceleration in prediction of a different RT. The mean error for conventional PINN and TL-PINN models prediction of neutron densities is smaller than 1%. We have developed a correlation between TL-PINN performance acceleration and similarity measure of RTs, which can be used as a guide for application of TL-PINNs.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

PowerModelsGAT-AI: Physics-Informed Graph Attention for Multi-System Power Flow With Continual Learning

Solving the alternating current power flow equations in real time is essential for secure grid operation, yet classical Newton–Raphson solvers can be slow under stressed conditions. Existing graph neural networks for power flow are typically trained on a single system and often degrade on different systems. We present PowerModelsGAT-AI, a physics-informed graph attention network that predicts bus voltages and generator injections. The model uses bus-type-aware masking to handle different bus types and balances multiple loss terms, including a power-mismatch penalty, using learned weights. We evaluate the model on 14 benchmark systems (4 to 6,470 buses) and train a unified model on 13 of these under contingency conditions with up to two branch outages, achieving an average normalized mean absolute error of 0.89% for voltage magnitudes and R 2 >0.99 for voltage angles. We also show continual learning: when adapting a base model to a new 1,354-bus system, standard fine-tuning causes severe forgetting with error increases exceeding 1000% on base systems, while our experience replay and elastic weight consolidation strategy keeps error increases below 2% and in some cases improves base-system performance. Interpretability analysis shows that learned attention weights correlate with physical branch parameters (susceptance: r=0.38 ; thermal limits: r=0.22 ), and feature importance analysis supports that the model captures established power flow relationships.

24 POWER TRANSMISSION AND DISTRIBUTION↗

The Role of Margin in Link Design and Optimization

Link analysis is a system engineering process in the design, development, and operation of communication systems and networks. Link models that are mathematical abstractions representing the useful signal power and the undesirable noise and attenuation effects (including weather effects if the signal path transverses through the atmosphere) that are integrated into the link budget calculation that provides the estimates of signal power and noise power at the receiver. Then the link margin is applied which attempts to counteract the fluctuations of the signal and noise power to ensure reliable data delivery from transmitter to receiver. (Link margin is dictated by the link margin policy or requirements.) A simple link budgeting approach assumes link parameters to be deterministic values typically adopted a rule-of-thumb policy of 3 dB link margin. This policy works for most S- and X-band links due to their insensitivity to weather effects. But for higher frequency links like Ka-band, Ku-band, and optical communication links, it is unclear if a 3 dB link margin would guarantee link closure. Statistical link analysis that adopted the 2-sigma or 3-sigma link margin incorporates link uncertainties in the sigma calculation. (The Deep Space Network (DSN) link margin policies are 2-sigma for downlink and 3-sigma for uplink.) The link reliability can therefore be quantified statistically even for higher frequency links. However in the current statistical link analysis approach, link reliability is only expressed as the likelihood of exceeding the signal-to-noise ratio (SNR) threshold that corresponds to a given bit-error-rate (BER) or frame-error-rate (FER) requirement. The method does not provide the true BER or FER estimate of the link with margin, or the required signalto-noise ratio (SNR) that would meet the BER or FER requirement in the statistical sense. In this paper, we perform in-depth analysis on the relationship between BER/FER requirement, operating SNR, and coding performance curve, in the case when the channel coherence time of link fluctuation is comparable or larger than the time duration of a codeword. We compute the "true" SNR design point that would meet the BER/FER requirement by taking into account the fluctuation of signal power and noise power at the receiver, and the shape of the coding performance curve. This analysis yields a number of valuable insights on the design choices of coding scheme and link margin for the reliable data delivery of a communication system - space and ground. We illustrate the aforementioned analysis using a number of standard NASA error-correcting codes.

Cheung, K.↗

Real-time Identification and Control of Satellite Signal Impairments Solution and Application of the Stratonovich Equation Part 1. Theoretical Development

As satellite communications systems become both more complex and reliant with respect to their operating environment, it has become imperative to be able to identify, during real-time operation, the onset of one or more impairments to the quality of overall communications system integrity. One of the most important aspects to monitor of a satellite link operating within the Earth's atmosphere is the signal fading due to the occurrence of rain and/or phase scintillations. This, of course, must be done in the presence of the associated measurement uncertainty or potentially faulty measurement equipment such as in the Advanced Communication Technology Satellite (ACTS) experiment. In the present work, an approach originally suggested in 1991, and apparently still considered iconoclastic, will be significantly developed and applied to the satellite communications link on which the deleterious composite signal fade is the result of one or many component fade mechanisms. Through the measurement (with the attendant uncertainty or 'error' in the measurement) of such a composite fading satellite signal, it is desired to extract the level of each of the individual fading mechanisms so they can be appropriately mitigated before they impact the overall performance of the communications network. Rather than employing simple-minded deterministic filtering to the real-time fading, the present approach is built around all the models and/or descriptions used to describe the individual fade components, including their dynamic evolution. The latter is usually given by a first-order Langevin equation. This circumstance allows the description of the associated temporal transition probability densities of each of the component processes. By using this description, along with the real-time measurements of the composite fade (along with the measurement errors), one can obtain statistical estimates of the levels of each of the component fading mechanisms as well as their predicted values into the future. This is all accomplished by the use of the well-known Stratonovich integro-differential equation that results from the model of the measured signal fade that is also tailored to adaptively adjust the values of the parameters used in the statistical models of the individual fade mechanisms. Three examples of increasing complexity are addressed and solved for the iterative determination of fade component levels from the measured composite signal fade in the presence of measurement error and, in the last case, with uncertainty in the model parameters.

Countermeasure↗

Encryption for Remote Control via Internet or Intranet

A data-communication protocol has been devised to enable secure, reliable remote control of processes and equipment via a collision-based network, while using minimal bandwidth and computation. The network could be the Internet or an intranet. Control is made secure by use of both a password and a dynamic key, which is sent transparently to a remote user by the controlled computer (that is, the computer, located at the site of the equipment or process to be controlled, that exerts direct control over the process). The protocol functions in the presence of network latency, overcomes errors caused by missed dynamic keys, and defeats attempts by unauthorized remote users to gain control. The protocol is not suitable for real-time control, but is well suited for applications in which control latencies up to about 0.5 second are acceptable. The encryption scheme involves the use of both a dynamic and a private key, without any additional overhead that would degrade performance. The dynamic key is embedded in the equipment- or process-monitor data packets sent out by the controlled computer: in other words, the dynamic key is a subset of the data in each such data packet. The controlled computer maintains a history of the last 3 to 5 data packets for use in decrypting incoming control commands. In addition, the controlled computer records a private key (password) that is given to the remote computer. The encrypted incoming command is permuted by both the dynamic and private key. A person who records the command data in a given packet for hostile purposes cannot use that packet after the public key expires (typically within 3 seconds). Even a person in possession of an unauthorized copy of the command/remote-display software cannot use that software in the absence of the password. The use of a dynamic key embedded in the outgoing data makes the central-processing unit overhead very small. The use of a National Instruments DataSocket(TradeMark) (or equivalent) protocol or the User Datagram Protocol makes it possible to obtain reasonably short response times: Typical response times in event-driven control, using packets sized .300 bytes, are <0.2 second for commands issued from locations anywhere on Earth. The protocol requires that control commands represent absolute values of controlled parameters (e.g., a specified temperature), as distinguished from changes in values of controlled parameters (e.g., a specified increment of temperature). Each command is issued three or more times to ensure delivery in crowded networks. The use of absolute-value commands prevents additional (redundant) commands from causing trouble. Because a remote controlling computer receives "talkback" in the form of data packets from the controlled computer, typically within a time interval < or =1 s, the controlling computer can re-issue a command if network failure has occurred. The controlled computer, the process or equipment that it controls, and any human operator(s) at the site of the controlled equipment or process should be equipped with safety measures to prevent damage to equipment or injury to humans. These features could be a combination of software, external hardware, and intervention by the human operator(s). The protocol is not fail-safe, but by adopting these safety measures as part of the protocol, one makes the protocol a robust means of controlling remote processes and equipment by use of typical office computers via intranets and/or the Internet.

Lineberger, Lewis↗

Modeling the Atmospheric Phase Effects of a Digital Antenna Array Communications System

In an antenna array system such as that used in the Deep Space Network (DSN) for satellite communication, it is often necessary to account for the effects due to the atmosphere. Typically, the atmosphere induces amplitude and phase fluctuations on the transmitted downlink signal that invalidate the assumed stationarity of the signal model. The degree to which these perturbations affect the stationarity of the model depends both on parameters of the atmosphere, including wind speed and turbulence strength, and on parameters of the communication system, such as the sampling rate used. In this article, we focus on modeling the atmospheric phase fluctuations in a digital antenna array communications system. Based on a continuous-time statistical model for the atmospheric phase effects, we show how to obtain a related discrete-time model based on sampling the continuous-time process. The effects of the nonstationarity of the resulting signal model are investigated using the sample matrix inversion (SMI) algorithm for minimum mean-squared error (MMSE) equalization of the received signal

Tkacenko, A.↗

Estimating Ground-Level PM(sub 2.5) Concentrations in the Southeastern United States Using MAIAC AOD Retrievals and a Two-Stage Model

Previous studies showed that fine particulate matter (PM(sub 2.5), particles smaller than 2.5 micrometers in aerodynamic diameter) is associated with various health outcomes. Ground in situ measurements of PM(sub 2.5) concentrations are considered to be the gold standard, but are time-consuming and costly. Satellite-retrieved aerosol optical depth (AOD) products have the potential to supplement the ground monitoring networks to provide spatiotemporally-resolved PM(sub 2.5) exposure estimates. However, the coarse resolutions (e.g., 10 km) of the satellite AOD products used in previous studies make it very difficult to estimate urban-scale PM(sub 2.5) characteristics that are crucial to population-based PM(sub 2.5) health effects research. In this paper, a new aerosol product with 1 km spatial resolution derived by the Multi-Angle Implementation of Atmospheric Correction (MAIAC) algorithm was examined using a two-stage spatial statistical model with meteorological fields (e.g., wind speed) and land use parameters (e.g., forest cover, road length, elevation, and point emissions) as ancillary variables to estimate daily mean PM(sub 2.5) concentrations. The study area is the southeastern U.S., and data for 2003 were collected from various sources. A cross validation approach was implemented for model validation. We obtained R(sup 2) of 0.83, mean prediction error (MPE) of 1.89 micrograms/cu m, and square root of the mean squared prediction errors (RMSPE) of 2.73 micrograms/cu m in model fitting, and R(sup 2) of 0.67, MPE of 2.54 micrograms/cu m, and RMSPE of 3.88 micrograms/cu m in cross validation. Both model fitting and cross validation indicate a good fit between the dependent variable and predictor variables. The results showed that 1 km spatial resolution MAIAC AOD can be used to estimate PM(sub 2.5) concentrations.

aerosol optical depth↗

Underlying Fundamentals of Kalman Filtering for River Network Modeling

The grand challenge of producing hydrometeorological estimates every time and everywhere has motivated the fusion of sparse observations with dense numerical models, with a particular interest on discharge in river modeling. Ensemble methods are largely preferred as they enable the estimation of error properties, but at the expense of computational load and generally with underestimations. These imperfect stochastic estimates motivate the use of correction methods, that is, error localization and inflation, although the physical justifications for their optimality are limited. The purpose of this study is to use one of the simplest forms of data assimilation when applied to river modeling and reveal the underlying mechanisms impacting its performance. Our framework based on assimilating daily averaged in situ discharge measurements to correct daily averaged runoff was tested over a 4-yr case study of two rivers in Texas. Results show that under optimal conditions of inflation and localization, discharge simulations are consistently improved such that the mean values of Nash–Sutcliffe efficiency are enhanced from211.32 to 0.55 at observed gauges and from212.24 to21.10 at validation gauges. Yet, parameters controlling the inflation and the localization have a large impact on the performance. Further investigations of these sensitivities showed that optimal inflation occurs when compensating exactly for discrepancies in the magnitude of errors while optimal localization matches the distance traveled during one assimilation window. These results may be applicable to more advanced data assimilation methods as well as for larger applications motivated by upcoming river-observing satellite missions, such as NASA’s Surface Water and Ocean Topography mission.

Streamflow↗

Automated assignment of rotational spectra using artificial neural networks

Here, a typical broadband rotational spectrum may contain several thousand observable transitions, spanning many species. While these spectra often encode troves of chemical information, identifying and assigning the individual spectra can be challenging. Traditional approaches typically involve visually identifying a pattern. A more modern approach is to apply an automated fitting routine. In this approach, combinations of 3 transitions are searched by trial and error, to fit the A , B , and C rotational constants in a Watson-type Hamiltonian. In this work, we develop an alternative approach-to utilize machine learning to train a computer to recognize the patterns inherent in rotational spectra. Broadband high-resolution rotational spectra are perhaps uniquely suited for pattern recognition, assignment, and species identification using machine learning. Repeating patterns of transition frequencies and intensities are now routinely recorded in broadband chirped-pulse Fourier transform microwave experiments in which both the number of resolution elements and the dynamic range surpass 10 4 . At the same time, these high-resolution spectra are extremely sensitive to molecular geometry with each polar species having a unique rotational spectrum. Here we train the feed forward neural network on thousands of rotational spectra that we calculate, using the rules of quantum mechanics, from randomly generated sets of rotational constants and other Hamiltonian parameters. Reasonable physical constraints are applied to these parameter sets, yet they need not belong to existing species. A trained neural network presented with a spectrum identifies its type (e.g., linear molecule, symmetric top, or asymmetric top) and infers the corresponding Hamiltonian parameters (rotational constants, distortion, and hyperfine constants). The classification and prediction times, about 160 μs and 50 μs, respectively, seem independent of the spectral complexity or the number of molecular parameters. We describe how the network works, provide benchmarking results, and discuss future directions.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Deep learning surrogate model for kinetic Landau-fluid closure with collision

In this work, the kinetic Landau-fluid (LF) closure with collision and periodic boundary condition is used in the development of the deep learning (DL) surrogate model. A classical neural network, namely, feedforward neural network or sometimes termed multilayer perceptron, is constructed and trained to learn the kinetic LF closure in the static limit and arbitrary mean free path in configuration space. The preliminary relation between best hyperparameters and critical parameters for data generation is found. Compared with the numerical approach (non-Fourier method) of the LF closure, the deep learning surrogate model shows an order of magnitude of improvement in terms of accuracy. Perhaps most importantly, the surrogate model closure has been integrated for the first time with fluid simulations. Our DL-enabled fluid simulations, for the first time, give the correct Landau damping rate for a wide range of wave vectors, while the Hammett–Perkins closure cannot produce the correct damping rate. We correctly connect the collisionless Hammett–Perkins closure and collisional Braginskii closure to reproduce the intrinsic nonlocal feature of the heat flux with DL techniques. We address the most concerning error accumulation problem and find that simulations with the deep learning surrogate model are as good as, if not better than, simulations with the analytic closure in terms of long-term numerical stability in the linear Landau damping test.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Accurate Compression of Tabulated Chemistry Models with Partition of Unity Networks

Tabulated chemistry models are widely used to simulate large-scale turbulent fires in applications including energy generation and fire safety. Tabulation via piecewise Cartesian interpolation suffers from the curse-of-dimensionality, leading to a prohibitive exponential growth in parameters and memory usage as more dimensions are considered. Artificial neural networks (ANNs) have attracted attention for constructing surrogates for chemistry models due to their ability to perform high-dimensional approximation. However, due to well-known pathologies regarding the realization of suboptimal local minima during training, in practice they do not converge and provide unreliable accuracy. Partition of unity networks (POUnets) are a recently introduced family of ANNs which preserve notions of convergence while performing high-dimensional approximation, discovering a mesh-free partition of space which may be used to perform optimal polynomial approximation. In this work, we assess their performance with respect to accuracy and model complexity in reconstructing unstructured flamelet data representative of nonadiabatic pool fire models. Our results show that POUnets can provide the desirable accuracy of classical spline-based interpolants with the low memory footprint of traditional ANNs while converging faster to significantly lower errors than ANNs. For example, we observe POUnets obtaining target accuracies in two dimensions with 40 to 50 times less memory and roughly double the compression in three dimensions. We also address the practical matter of efficiently training accurate POUnets by studying convergence over key hyperparameters, the impact of partition/basis formulation, and the sensitivity to initialization.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Automatically generated model for light alkene combustion

Light alkenes are common combustion intermediates for a variety of fuels. Therefore, understanding their oxidation and pyrolysis chemistry is key to building detailed mechanisms for heavier fuels. This work was focused on the development and evaluation of a detailed kinetic mechanism suitable for the combustion of light alkenes up to C 4 without the use of tuned parameters, instead, the parameter values come from first principles or direct measurements. The generated mechanism accurately estimates the laminar burning velocity (S u ) and ignition delay time (IDT) of light alkenes available in the literature, which represent fundamental combustion properties at a wide range of conditions. Because each parameter is thought to have a physically realistic value, not tuned to these measurements, the new model could be used as a sub-mechanism in models for other applications. The reaction network was generated with the open-source Reaction Mechanism Generator (RMG) software. Sensitivity analyses were performed under wide ranges of temperatures and pressures, allowing for the identification of the most impactful species and reactions. Based on these, a comprehensive thermochemistry database, including calculations on 550 molecules performed in this work at the CBS-QB3 level of theory, and a kinetic library, including theoretically-derived reaction rates retrieved from the literature, were built and used in the mechanism generation. The developed mechanism was compared against several existing detailed kinetic mechanisms for ethene, propene, 1-butene, 2-butene, and isobutene. Here, the newly generated model is the most accurate among the ones analyzed, in terms of fractional bias and normalized mean square error. Hence, this new model was used to analyze the chemistry of alkene combustion. Key rate coefficients were compared, to identify the cause of deviations between the models and possible areas for further improvements.

09 BIOMASS FUELS↗