Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Performance Tuning”

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 649 records · Page 36
All sourcesNASA NTRSDOE OSTI
DOE OSTI2023

Building thermal dynamics modeling with deep transfer learning using a large residential smart thermostat dataset

Understanding thermal dynamics and obtaining the computational model of residential buildings enable its scaled application in energy retrofits, control optimization and decarbonization. In this paper, we present a deep learning approach to model building thermal dynamics with smart thermostat data collected from residential buildings, with the goal to investigate model generalizability. In the first stage, we developed and compared different Deep Learning architectures including Convolutional Neural Networks (CNN), Long Short-Term Memory (LSTM) models and CNN-LSTM to predict indoor air temperature in a multi-step time horizon. In the second stage, we implemented a Transfer Learning (TL) process, which aims to improve the prediction performance on a new set of buildings (targets), exploiting the knowledge of related or similar buildings (sources). Different TL strategies and source model identification methods were investigated. The study showed that the CNN-LSTM performed the best among the architectures compared, with an average Mean Absolute Error (MAE) of 0.26 °C for one-hour-ahead (twelve 5-min future steps) predictions. Furthermore, the results showed that freezing the LSTM layer and fine-tuning the other layers of the CNN-LSTM achieved the best performance among four TL strategies, which further improved the performance with respect to a machine learning approach by 10%, and proving the effectiveness and generalizability of the proposed approach. A comparison of three different source model identification methods showed that randomly selecting source models constrained by similar building characteristics can provide good TL performance while retaining simplicity comparing with other quantitative source identification methods.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗
DOE OSTI2023

Operational control with set-points tuning - application to mobile robots

This paper proposes a novel method for the optimal tuning of set points for multiple-layered control system structure widely seen in robotics and other complex industrial processes composed of a number of subsystems. The terminal sliding mode control (SMC) is used as the low-level control strategy to ensure the stability of subsystems. When uncertainties exist, it can be shown that the deteriorated system performance will be improved by the outer loop with set points tuning. For this purpose, the learning of the new set point is designed to compensate for the effects caused by uncertainties during the system operation. At the same time, the system is proven to stay with the original set point when the compensation is introduced. A practical application to a holonomic mobile robot system is given to illustrate the presented method. Desired results have been obtained.

42 ENGINEERING↗
DOE OSTI2020

Compiler and Runtime Approaches to Enable Large-Scale Irregular Programs. Final report, July 2013 - July 2019

While regular algorithms, characterized by operations on dense matrices and arrays, have long been the mainstay of scientific, high-performance computing, irregular algorithms, which feature unpredictable accesses to pointer-based data structures, are becoming increasingly common in high performance computing, arising in graph analysis, data mining and visualization, among other domains. Unfortunately, the defining characteristics of irregular applications, their dynamic, unpredictable, data-dependent access patterns and data layouts, make achieving high performance on large scale systems difficult. Scaling applications to peta- and exa-scale requires carefully controlling communication and data movement and placement, an inherently difficult task when access patterns and data layouts are unpredictable! Most irregular applications that attain high performance must be painstakingly hand-written and hand-tuned, with few common principles or paradigms uniting various implementations and easing future development. Despite the increasing importance of irregular applications, there is little programmer knowledge, and even less compiler ability, devoted to optimizing them. This project aims to solve these problems. By allowing programmers to write irregular applications in high level forms, with at most a few annotations highlighting key structural properties, programmers can focus on developing their algorithms and methods. The compiler and run-time system can take on the tedious task of optimizing the application for execution at large scales, and can automatically provide efficient implementations. This will provide portability and ease maintenance for existing irregular applications, but, more importantly, open up whole new domains of computational science to large-scale, high-performance simulation codes.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Robust wind farm layout optimization

Wake interactions in wind farms cause losses in annual energy production (AEP) on the order of 10%. Wind farm designers optimize the layout of the farm to mitigate wake losses, especially in the dominant site-specific wind directions. As wind turbines and wind farms grow in scale, optimization becomes more complex. Offshore wind farms regularly comprise more than 100 wind turbines and are characterized by complex boundaries due to shipping lanes, neighboring wind farms, and other constraints. Layout optimization methods are broadly split between gradient-based and gradient-free approaches. Gradient-based approaches can converge quickly and perform well for smaller, academic problems but are often sensitive to initial conditions and tuning parameters and require expert knowledge to use. On the other hand, gradient-free approaches can be more robust to problem complexities. We present a robust layout optimization approach based on a random search algorithm. The algorithm is intended for those who are not optimization experts and has few tuning parameters that need specification to achieve satisfactory results. Unlike off-the-shelf methods, which use generally available, non-domain-specific optimization routines that accept as inputs an optimization function and constraint definitions, this approach takes advantage of the relative computational costs of the different evaluations by evaluating cheaper computations first (boundary and minimum distance constraints) and running expensive AEP evaluations only if all other checks pass. Moreover, an outer genetic algorithm allows multiple solutions to evolve in parallel, enabling rapid solution development on high-performance computers. We discuss the relative ease of selecting necessary tuning parameters and demonstrate the efficacy of the genetic random search on a complex layout problem consisting of placing 70 turbines in a nonconvex and unconnected boundary region.

17 WIND ENERGY↗
DOE OSTI2021

High-Performance Catalytic Four-Channel Hollow Fibers with Highly Dispersed Nickel Nanoparticles Prepared by Atomic Layer Deposition for Dry Reforming of Methane

In this study, highly dispersed nickel (Ni) nanoparticles (NPs) with an average particle size of 4.3 nm were uniformly deposited on the outer surface, the inner channel surface, and inside the pores of 20-cm long 4-channel α-Al 2 O 3 hollow fibers (HFs) by atomic layer deposition (ALD) for dry reforming of methane (DRM). Cerium oxide (CeO 2 ) was added to promote the catalytic performance of Ni/Al 2 O 3 -HF catalysts. Rationally designed filling methods, by tuning the reactor size and inert fillings, can reduce the catalyst bed voidage in a fixed bed reactor for better reactant gas distribution, effectively utilize the Ni reactive sites, and achieve excellent catalytic performance. It was found that the CeO 2 -promoted Ni/Al 2 O 3 -HF catalyst was highly active and highly stable without deactivation during an overall 400-hr DRM test at 850 ºC. CeO 2 with reversible valence states could participate in surface reactions; especially, the formation of CeAlO 3 provided sufficient surface Ce 3+ for CO 2 activation and enhanced the stability and reusability of the HF catalysts.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2022

Implementation of a Binary Neural Network on a Passive Array of Magnetic Tunnel Junctions

The increasing scale of neural networks and their growing application space have produced demand for more energy- and memory-efficient artificial-intelligence-specific hardware. Avenues to mitigate the main issue, the von Neumann bottleneck, include in-memory and near-memory architectures, as well as algorithmic approaches. In this report we leverage the low-power and the inherently binary operation of magnetic tunnel junctions (MTJs) to demonstrate neural network hardware inference based on passive arrays of MTJs. In general, transferring a trained network model to hardware for inference is confronted by degradation in performance due to device-to-device variations, write errors, parasitic resistance, and nonidealities in the substrate. To quantify the effect of these hardware realities, we benchmark 300 unique weight matrix solutions of a two-layer perceptron to classify the Wine dataset for both classification accuracy and write fidelity. Despite device imperfections, we achieve software-equivalent accuracy of up to 95.3% with proper tuning of network parameters in 15 x 15 MTJ arrays having a range of device sizes. The success of this tuning process shows that new metrics are needed to characterize the performance and quality of networks reproduced in mixed signal hardware.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
NASA NTRS2021

A Learn-to-Fly Approach for Adaptively Tuning Flight Control Systems

A method is presented for adaptively tuning feedback control gains in a flight control system to achieve desired closed-loop performance. The method combines efficient aircraft parameter estimation, for identifying closed-loop dynamics models, with online nonlinear optimization, for sequentially perturbing and updating control gains to improve performance. Access to flight measurements and the control gains is required, but no prior information about the aircraft dynamics or the flight control architecture is needed. After the procedure, the optimized control gains (with uncertainties), the open-loop dynamics model, and the closed-loop dynamics model are available. The method is demonstrated for tuning a longitudinal stability augmentation system using a realistic nonlinear flight dynamics simulation of a subscale airplane. Convergence was attained using five maneuvers that spanned approximately one minute of flight test time. Although shown for a relatively simple case, the method is general and can be applied to other aircraft, axes of motion, performance metrics, and control system designs.

Adaptive tuning↗
DOE OSTI2026

Heterogeneous Multilayer Nanopores via Chemically Tuned Dielectric Breakdown for Single‐Molecule Sensing

Solid-state nanopores are powerful platforms for single-molecule sensing, yet their performance is often constrained by fabrication complexity, noise, and limited control over surface properties. Here we report a direct method to fabricate heterogeneous multilayer nanopores using chemically tuned controlled dielectric breakdown (CT-CDB). We integrate hBN, MoS 2 , or graphene atop a silicon nitride membrane to form five distinct bilayer and tri-layer architectures, with bare SiN x nanopore as a control. CT-CDB achieves pore formation reproducibly through material-stacks with high efficiency, good pore size control, and strong yield, validated by various characterizations. Transferrin protein translocation experiments, supported by simulations, reveal that multilayer configurations modulate protein conformations, ionic current blockade and dwell time distributions, reflecting combined effects of membrane type, interfacial chemistry, and local electric field gradients. A supervised machine learning framework is implemented to assist identifying multilayer structure effects embedded in signal signatures, with over 96% accuracy. This work presents a modular and scalable framework for functional nanopore engineering with complex structural integration, thereby expanding the potential of 2D materials in single-molecule sensing applications.

2D materials↗
NASA NTRS1990

An expert system to perform on-line controller restructuring for abrupt model changes

Work in progress on an expert system used to reconfigure and tune airframe/engine control systems on-line in real time in response to battle damage or structural failures is presented. The closed loop system is monitored constantly for changes in structure and performance, the detection of which prompts the expert system to choose and apply a particular control restructuring algorithm based on the type and severity of the damage. Each algorithm is designed to handle specific types of failures and each is applicable only in certain situations. The expert system uses information about the system model to identify the failure and to select the technique best suited to compensate for it. A depth-first search is used to find a solution. Once a new controller is designed and implemented it must be tuned to recover the original closed-loop handling qualities and responsiveness from the degraded system. Ideally, the pilot should not be able to tell the difference between the original and redesigned systems. The key is that the system must have inherent redundancy so that degraded or missing capabilities can be restored by creative use of alternate functionalities. With enough redundancy in the control system, minor battle damage affecting individual control surfaces or actuators, compressor efficiency, etc., can be compensated for such that the closed-loop performance in not noticeably altered. The work is applied to a Black Hawk/T700 system.

Litt, Jonathan S.↗
DOE OSTI2021

Machine-learning-accelerated multimodal characterization and multiobjective design optimization of natural porous materials

Natural porous materials such as nanoporous clays are used as green and low-cost adsorbents and catalysts. The key factors determining their performance in these applications are the pore morphology and surface activity, which are typically represented by properties such as specific surface area, pore volume, micropore content and pH. The latter may be modified and tuned to specific applications through material processing and/or chemical treatment. Characterization of the material, raw or processed, is typically performed experimentally, which can become costly especially in the context of tuning of the properties towards specific application requirements and needing numerous experiments. In this work, we present an application of tree-based machine learning methods trained on experimental datasets to accelerate the characterization of natural porous materials. The resulting models allow reliable prediction of the outcomes of experimental characterization of processed materials (R2 from 0.78 to 0.99) as well as identification of key factors contributing to those properties through feature importance analysis. Furthermore, the high throughput of the models enables exploration of processing parameter–property correlations and multiobjective optimization of prototype materials towards specific applications. We have applied these methodologies to pinpoint and rationalize optimal processing conditions for clays exploitable in acid catalysis. One of such identified materials was synthesized and tested revealing appreciable acid character improvement with respect to the pristine material. Specifically, it achieved 79% removal of chlorophyll-a in acid catalyzed degradation.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2025

A machine-learning-driven data labeling pipeline for scientific analysis in MLExchange

This study introduces a novel labeling pipeline to accelerate the labeling process of scientific data sets by using artificial intelligence (AI)-guided tagging techniques. This pipeline includes a set of interconnected web-based graphical user interfaces (GUIs), where Data Clinic and MLCoach enable the preparation of machine learning (ML) models for data reduction and classification, respectively, while Label Maker is used for label assignment. Throughout this pipeline, data can be accessed through a direct connection to a file system or through Tiled for access through Hypertext Transfer Protocol (HTTP). Our experimental results present three use cases where this labeling pipeline has been instrumental for the study of large X-ray scattering data sets in the area of pattern recognition, the remote analysis of resonant soft X-ray scattering data and the fine-tuning process of foundation models. These use cases highlight the labeling capabilities of this pipeline, including the ability to label large data sets in a short period of time, to perform remote data analysis while minimizing data movement and to enhance the fine-tuning process of complex ML models with human involvement.

Chavez, Tanny (ORCID:0000000193172896)↗
DOE OSTI2023

KEBLM: Knowledge-Enhanced Biomedical Language Models

Pretrained language models (PLMs) have demonstrated strong performance on many natural language processing (NLP) tasks. Despite their great success, these PLMs are typically pretrained only on unstructured free texts without leveraging existing structured knowledge bases that are readily available for many domains, especially scientific domains. As a result, these PLMs may not achieve satisfactory performance on knowledge-intensive tasks such as biomedical NLP. Comprehending a complex biomedical document without domain-specific knowledge is challenging, even for humans. Inspired by this observation, we propose a general framework for incorporating various types of domain knowledge from multiple sources into biomedical PLMs. We encode domain knowledge using lightweight adapter modules, bottleneck feed-forward networks that are inserted into different locations of a backbone PLM. For each knowledge source of interest, we pretrain an adapter module to capture the knowledge in a self-supervised way. We design a wide range of self-supervised objectives to accommodate diverse types of knowledge, ranging from entity relations to description sentences. Once a set of pretrained adapters is available, we employ fusion layers to combine the knowledge encoded within these adapters for downstream tasks. Each fusion layer is a parameterized mixer of the available trained adapters that can identify and activate the most useful adapters for a given input. Our method diverges from prior work by including a knowledge consolidation phase, during which we teach the fusion layers to effectively combine knowledge from both the original PLM and newly-acquired external knowledge using a large collection of unannotated texts. After the consolidation phase, the complete knowledge-enhanced model can be fine-tuned for any downstream task of interest to achieve optimal performance. Extensive experiments on many biomedical NLP datasets show that our proposed framework consistently improves the performance of the underlying PLMs on various downstream tasks such as natural language inference, question answering, and entity linking. These results demonstrate the benefits of using multiple sources of external knowledge to enhance PLMs and the effectiveness of the framework for incorporating knowledge into PLMs. Finally, while primarily focused on the biomedical domain in this work, our framework is highly adaptable and can be easily applied to other domains, such as the bioenergy sector.

96 KNOWLEDGE MANAGEMENT AND PRESERVATION↗
DOE OSTI2022

Tuning Fluorination of Linear Carbonate for Lithium-Ion Batteries

Liquid electrolyte engineering plays a critical role in modern lithium-ion batteries. However, the existing electrolytes fall short when used with some trending battery chemistries such as high-voltage and high-energy-density electrodes. Fluorination of electrolyte solvents has been identified as an effective approach for improved cyclability, but few works systematically studied the effects of fluorination extent of carbonate solvents on battery performance. Here we design and synthesize a family of fluorinated ethyl methyl carbonates. Different numbers of F atoms are finely tuned to yield monofluoroethyl methyl carbonate (F1EMC), difluoroethyl methyl carbonate (F2EMC) and trifluoroethyl methyl carbonate (F3EMC). The cycling behavior of several types of lithium-ion pouch cells, including graphite (Gr)/single-crystalline LiNi 0.8 Mn 0.1 Co 0.1 O 2 (SC-NMC811), Gr-SiO x /LiNi 0.6 Mn 0.2 Co 0.2 O 2 (NMC622), high-voltage Gr/LiNi 0.5 Mn 1.5 O 4 (LNMO), Gr/layered Li-rich Mn-based oxide (LLMO) and fast-charging Gr/NMC622, were systematically investigated to understand the impact of fluorination degree. Compared to the commercially available F3EMC, we found that the partially-fluorinated F1EMC and F2EMC in some cases showed improved cycling stability, which we attribute to their locally-polar –CH 2 F and –CHF 2 groups and thus fast ion conduction than –CF 3 . This work suggests that highly or fully fluorinated solvents are not necessarily desirable; instead, fluorination degree needs to be rationally and finely tuned for optimized lithium-ion cell performance.

25 ENERGY STORAGE↗
DOE OSTI2022

A reference open-source controller for fixed and floating offshore wind turbines

Abstract. This paper describes the development of a new reference controller framework for fixed and floating offshore wind turbines that greatly facilitates controller tuning and represents standard industry practices. The reference wind turbine controllers that are most commonly cited in the literature have been developed to work with specific reference wind turbines. Although these controllers have provided standard control functionalities, they are often not easy to modify for use on other turbines, so it has been challenging for researchers to run representative, fully dynamic simulations of other wind turbine designs. The Reference Open-Source Controller (ROSCO) has been developed to provide a modular reference wind turbine controller that represents industry standards and performs comparably to or better than existing reference controllers. The formulation of the ROSCO controller logic and tuning processes is presented in this paper. Control capabilities such as tip speed ratio tracking generator torque control, minimum pitch saturation, wind speed estimation, and a smoothing algorithm at near-rated operation are included to provide modern controller features. A floating offshore wind turbine feedback module is also included to facilitate growing research in the floating offshore arena. All of the standard controller implementations and control modules are automatically tuned such that a non-controls engineer or automated optimization routine can easily improve the controller performance. This article provides the framework and theoretical basis for the ROSCO controller modules and generic tuning processes. Simulations of the National Renewable Energy Laboratory (NREL) 5 MW reference wind turbine and International Energy Agency 15 MW reference turbine on the University of Maine semisubmersible platform are analyzed to demonstrate the controller's performance in both fixed and floating configurations, respectively. The simulation results demonstrate ROSCO's peak shaving routine to reduce maximum rotor thrusts by over 10 % compared to the NREL 5 MW reference wind turbine controller on the land-based turbine and to reduce maximum platform pitch angles by nearly 30 % when using the platform feedback routine instead of a more traditional low-bandwidth controller.

17 WIND ENERGY↗
DOE OSTI2025

Sub‐5 Ångstrom Porosity Tuning in Calixarene‐Derived Porous Liquids via Supramolecular Complexation Construction

Abstract Sub‐Ångstrom‐level porosity engineering, which is appealing in gas separations, has been demonstrated in solid carbon, polymer, and framework materials but rarely achieved in the liquid phase. In this work, a gas molecular sieving effect in the liquid phase at sub‐5 Ångstrom scale is created via sophisticated porosity tuning in calixarene‐derived porous liquids (PLs). Type II PLs are constructed via supramolecular complexation between the sodium salts of calixarene derivatives and crown ether solvents. The chemical structure variation and assembly behavior of the porous host upon PL construction are monitored by spectroscopy‐, X‐ray‐, and neutron‐scattering techniques. The presence of permanent porosity in calixarene‐derived PLs is verified by pressure swing gas uptake, altered CO 2 physisorption behavior, and molecular simulations. Sub‐5 Ångstrom porosity tuning within the PL phase is achieved by introducing bulky substituted groups on the benzene ring of the calixarene host, which then greatly affects the dynamic motion and transport behavior of CO 2 molecules and the Xe uptake performance. The approach being demonstrated in this work represents a promising pathway to tune and leverage the porosity effect for enhanced gas uptake capacity and selectivity in liquid sorbents.

Li, Errui [Department of Chemistry University of T↗
DOE OSTI2022

Final Scientific/Technical Report-FREE-piston Zero Emissions Refrigerator (FREEZER)

The FREE-piston Zero Emissions Refrigerator (FREEZER™) is a natural refrigerant Stirling-cycle cooler developed by AMSC in partnership with Heatcraft, targeted as a replacement refrigeration system for walk-in coolers or freezers found commonly in restaurants, convenience stores, supermarkets, etc. The FREEZER™, combined with a closed-loop CO2 heat-transfer system, had the goal to provides an efficient (Target COP = 1.4), zero ODP (Ozone Depletion Potential) and near zero GWP (Global Warming Potential) refrigeration solution with no synthetic, flammable, or toxic refrigerants. Other near zero GWP refrigeration systems do not exist at the scale and performance level of FREEZER. ARPA-E Continuation Funding enabled the design and fabrication of one Advanced Development Unit (ADU) cooler and two Field Test Unit (FTU) coolers, and a new revised Field Test Unit (new FTU). Heatcraft was to extensively test the various coolers at their facilities, which closely simulate real-world conditions. The original plan was to install and operate FREEZER FTUs alongside a traditional cooling solution at one or more customer locations to obtain data on direct performance comparisons with existing commercial refrigeration systems. Success in this FREEZER refinement and field-test project will establish the basis for a new class of environmentally friendly refrigeration systems that will revolutionize the industry. The hardware developed under the program is shown below in Figure 1. Five Stirling-cycle machines were built within the program, with a sixth cooler still under construction. The original DU and the ADU coolers met the performance target with a coefficient of performance (COP) of 1.2. The FTU was underperforming during the initial tuning of the free-piston Stirling cycle dynamics when there was a catastrophic failure of the fasteners on pressure boundary. This caused a pause to the program while the failure mechanism was investigated. The root cause was an inadequate flange design that allowed an excessive cyclic loading on the fasteners, leading to a fatigue failure. It was determined that existing FTU design could not be reasonably salvaged in a safe manner, and that fact, combined with the lower-than-expected performance of the new experimental heat exchanger modules lead to the design of the “new FTU” with a more conventional shell-and-tube heat exchanger design which should get close to the design target COP of 1.4, but have a higher overall machine cost. The program ran out of time before the fabrication of the new FTU was completed, but AMSC is still working on the cooler, with the majority of the parts fabricated, and hopes to have the unit under test sometime in early 2022, after which it will be shipped to Heatcraft for additional testing.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗
DOE OSTI2025

GCAM Regional Tuning: A framework to tune GCAM parameters

GCAM assumptions typically generate scenarios that are designed to be internally consistent and globally coherent. The gcamdata tool which facilitates the compilation of data sets and user assumptions is not well suited to tailoring to specific country or regional realities, sponsor requirements, or perform harmonization for model intercomparison needs. As described in this report, the GCAM Regional Tuning project develops a computational framework that enables users to adjust GCAM parameters, so model outputs match targeted outcomes at user-defined spatial, temporal, and sectoral resolutions. The framework integrates GCAM, gcamdata, and gcamwrapper with a set of flexible “tuning directives” and an iterative numerical solver. Users can define targets (e.g., technology shares in power generation, BEV uptake, sectoral service demands), select tuners that manipulate relevant GCAM parameters (e.g., share weights, cost adders, elasticities), and export tuned parameters as reusable GCAM XML inputs for future runs. We demonstrate the approach and document usage, diagnostics, and known limitations, and we outline potential future directions.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

MechBERT: Language Models for Extracting Chemical and Property Relationships about Mechanical Stress and Strain

Language models are transforming materials-aware naturallanguage processing by enabling the extraction of dynamic, context-rich information from unstructured text, thus, moving beyond the limitations of traditional information-extraction methods. Moreover, small language models are on the rise because some of them can perform better than large language models (LLMs) when given domain-specific questionanswer tasks, especially about an application area that relies on a highly specialized vernacular, such as materials science. We therefore present a new class of MechBERT language models for understanding mechanical stress and strain in materials. These employ Bidirectional Encoder Representations for transformer (BERT) architectures. We showcase four MechBERT models, all of which were pretrained on a corpus of documents that are textually rich in chemicals and their stress–strain properties and were fine-tuned on question-answering tasks. We evaluated the level of performance of our models on domain-specific as well as general English-language question-answer tasks and also explored the influence of the size and type of BERT architectures on model performance. We find that our MechBERT models outperform BERT-based models of the same size and maintain relevancy better than much larger BERT-based models when tasked with domain-specific question-answering tasks within the stress–strain engineering sector. These small language models also enable much faster processing and require a much smaller fraction of data to pretrain them, affording them greater operational efficiency and energy sustainability than LLMs.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗