scholarly journals A Deep Learning Modeling Framework to Capture Mixing Patterns in Reactive-Transport Systems

2022 ◽  
Vol 31 (1) ◽  
pp. 188-223
Author(s):  
N. V. Jagtap
Keyword(s):  
Deep Learning ◽  
Reactive Transport ◽  
Transport Systems ◽  
Modeling Framework ◽  
Mixing Patterns

scholarly journals A hydro-thermal-geochemical modeling framework to simulate reactive transport in a waste coal area under amended and non-amended conditions

Heliyon ◽  
2020 ◽  
Vol 6 (1) ◽  
pp. e02803 ◽  
Author(s):  
Yi Xu ◽  
Fernando J. Plaza ◽  
Xu Liang ◽  
Tyler W. Davis ◽  
Judodine Nichols ◽  
...  
Keyword(s):  
Reactive Transport ◽  
Geochemical Modeling ◽  
Modeling Framework ◽  
Waste Coal

Deep Learning System for Travel Speed Predictions on Multiple Arterial Road Segments

2019 ◽  
Vol 2673 (4) ◽  
pp. 145-157 ◽  
Author(s):  
Hoang Nguyen ◽  
Christopher Bentley ◽  
Le Minh Kieu ◽  
Yushuai Fu ◽  
Chen Cai
Keyword(s):  
Deep Learning ◽  
Travel Speed ◽  
Learning System ◽  
Transport Systems ◽  
Practical Implementation ◽  
Generation Model ◽  
Vehicle Speed ◽  
Learning Methods ◽  
Arterial Road ◽  
Speed Prediction

Accurate travel speed prediction is a critical tool for incidence response management. The complex dynamics of transport systems render model-based prediction extremely challenging. However, the large amounts of available vehicle speed data contain the complex interdependencies of the target travel speed; the data itself can be used to generate accurate predictions using deep learning methods. In this work, a deep learning methodology involving feature generation, model development, and model deployment is presented. The authors demonstrate the high performance of deep learning methods (relative to more traditional benchmarks) in predicting travel speeds from 5–30 min in advance, for a challenging arterial road network. In this study, different deep learning architectures that exploit both spatial and temporal information for several time frames are compared and analyzed. Finally, the authors demonstrate the integration of their deep learning method into a visualization system that can be directly applied for vehicle speed prediction in real time. The model-selection analysis and data-to-visualization framework in this manuscript provide a step towards decision support for incident management; for practical implementation, the predictive power of deep learning models under incident conditions should continue to be investigated and improved.

White-Box and Black-Box Reliability Modeling Framework: Integration Through Analytical Model and User Profile Validation via Deep Learning — A Practitioner’s Approach

2021 ◽  
pp. 2140007
Author(s):  
K. Krishna Mohan ◽  
Harun Ul Rasheed Shaik ◽  
A. Srividya ◽  
Ajit Kumar Verma
Keyword(s):  
Deep Learning ◽  
Software Reliability ◽  
Reliability Assessment ◽  
User Profile ◽  
Black Box ◽  
Reliability Estimation ◽  
Box Model ◽  
Model Framework ◽  
Modeling Framework ◽  
Overall Reliability

Software reliability evaluation of complex systems is always a challenging task with conventional methods comprising both functional as well as nonfunctional aspects of real-world applications. Prevailing model frameworks moreover apply a nonfunctional approach (black-box model) that is modeled on defect data or through a functional approach (white-box model) that uses component or state-based interactions. Also, other challenges involve integrating both approaches, and validating user profiles of software operation. Further, reliability assessment is one among the most important and desirable qualities of service requirements of software systems, particularly in monitoring critical business transactions. Here, we propose a model framework to evaluate the overall reliability estimation involving both functional and nonfunctional model analyses using: (a) white-box assessment based on intercomponent analysis via component-based Cheung’s model and user profile validations with one of the identified deep learning techniques and (b) black-box modeling evaluation via generalized stochastic Petri nets based on orthogonal defect classification. A newly introduced deep learning model using white-box analysis is validated with pertinent usage profiles to establish a new trend in artificial neural networks and as well with software reliability estimation. Additionally, we introduce and present a quantitative technique — analytical hierarchy — to integrate reliability assessment and provide weights to the white-box and as well for black-box approaches to quantify overall reliability estimation. The proposed framework is illustrated with an application case study.

Pore-scale hydrodynamic evolution within carbonate rock during CO2 injection and sequestration

2021 ◽  
Author(s):  
Chi Zhang ◽  
Siyan Liu ◽  
Reza Barati
Keyword(s):  
High Resolution ◽  
Reactive Transport ◽  
Injection Rate ◽  
Mineral Dissolution ◽  
Co2 Injection ◽  
Pore Scale ◽  
Modeling Framework ◽  
Calcite Dissolution ◽  
Potential Factors ◽  
The Impact

<p><span>The continuously rising threat of global warming caused by human activities related to CO</span><span><sub>2</sub> emission is facilitating the development of greenhouse gas control technologies. Subsurface CO</span><span><sub>2</sub> injection and sequestration is one of the promising techniques to store CO</span><span><sub>2</sub> in the subsurface. </span><span> </span><span>However, during CO<sub>2</sub> injection, the mechanisms of processes like injectant immobilizations and trapping and pore-scale geochemical reactions such as mineral dissolution/precipitation are not well understood. Consequently, the multi-physics modeling approach is essential to elucidate the impact of all potential factors during CO<sub>2</sub> injection, thus to facilitate the optimization of this engineered application.</span> </p><p><span>Here, we propose a coupled framework to fully utilize the capabilities of the geochemical reaction solver PHREEQC while preserving the Lattice-Boltzmann Method (LBM) high-resolution pore-scale fluid flow integrated with diffusion processes. The model can simulate the dynamic fluid-solid interactions with equilibrium, kinetics, and surface reactions under the reactive-transport scheme.  In a simplified 2D spherical pack, we focused on examining the impact of pore sizes, grain size distributions, porosity, and permeability on the calcite dissolution/precipitation rate. Our simulation results show that the higher permeability, injection rate, and more local pore connectivity would significantly increase the reaction rate, then accelerate the pore-scale geometrical evolutions. Meanwhile, model accuracy is not sacrificed by reducing the number of reactants/species within the system.</span></p><p><span>Our modeling framework provides high-resolution details of the pore-scale fluid-solid interaction dynamics. To gain more insights into the mineral-fluid interfacial properties during CO</span><span><sub>2</sub> sequestration, our next step is to combine the electrodynamic forces into the model. Potentially, the proposed framework can be used for model upscaling and adaptive subsurface management in the future. </span><span> </span></p>

scholarly journals An explainable self-attention deep neural network for detecting mild cognitive impairment using multi-inputbdigital drawing tasks

2021 ◽  
Author(s):  
Natthanan Ruengchaijatuporn ◽  
Itthi Chatnuntawech ◽  
Surat Teerapittayanon ◽  
Sira Sriswasdi ◽  
Sirawaj Itthipuripat ◽  
...  
Keyword(s):  
Neural Network ◽  
Deep Learning ◽  
Cognitive Impairment ◽  
Mild Cognitive Impairment ◽  
Healthy Aging ◽  
Neurodegenerative Disorder ◽  
Early Stage ◽  
Aging Population ◽  
Modeling Framework ◽  
Drawing Tasks

Mild cognitive impairment (MCI) is an early stage of age-inappropriate cognitive decline, which could develop into dementia – an untreatable neurodegenerative disorder. An early detection of MCI is a crucial step for timely prevention and intervention. To tackle this problem, recent studies have developed deep learning models to detect MCI and various types of dementia using data obtained from the classic clock-drawing test (CDT), a popular neuropsychological screening tool that can be easily and rapidly implemented for assessing cognitive impairments in an aging population. While these models succeed at distinguishing severe forms of dementia, it is still difficult to predict the early stage of the disease using the CDT data alone. Also, the state-of-the-art deep learning techniques still face the black-box challenges, making it questionable to implement them in the clinical setting. Here, we propose a novel deep learning modeling framework that incorporates data from multiple drawing tasks including the CDT, cube-copying, and trail-making tasks obtained from a digital platform. Using self-attention and soft-label methods, our model achieves much higher classification performance at detecting MCI compared to those of a well-established convolutional neural network model. Moreover, our model can highlight features of the MCI data that considerably deviate from those of the healthy aging population, offering accurate predictions for detecting MCI along with visual explanation that aids the interpretation of the deep learning model.

scholarly journals 2D-to-3D image translation of complex nanoporous volumes using generative networks

2021 ◽  
Vol 11 (1) ◽  
Author(s):  
Timothy I. Anderson ◽  
Bolivia Vega ◽  
Jesse McKinzie ◽  
Saman A. Aryana ◽  
Anthony R. Kovscek
Keyword(s):  
Deep Learning ◽  
Reactive Transport ◽  
Ion Beam ◽  
Training Data ◽  
Imaging Modalities ◽  
High Contrast ◽  
Imaging Data ◽  
Contrast Imaging ◽  
Apparent Permeability ◽  
Image Translation

AbstractImage-based characterization offers a powerful approach to studying geological porous media at the nanoscale and images are critical to understanding reactive transport mechanisms in reservoirs relevant to energy and sustainability technologies such as carbon sequestration, subsurface hydrogen storage, and natural gas recovery. Nanoimaging presents a trade off, however, between higher-contrast sample-destructive and lower-contrast sample-preserving imaging modalities. Furthermore, high-contrast imaging modalities often acquire only 2D images, while 3D volumes are needed to characterize fully a source rock sample. In this work, we present deep learning image translation models to predict high-contrast focused ion beam-scanning electron microscopy (FIB-SEM) image volumes from transmission X-ray microscopy (TXM) images when only 2D paired training data is available. We introduce a regularization method for improving 3D volume generation from 2D-to-2D deep learning image models and apply this approach to translate 3D TXM volumes to FIB-SEM fidelity. We then segment a predicted FIB-SEM volume into a flow simulation domain and calculate the sample apparent permeability using a lattice Boltzmann method (LBM) technique. Results show that our image translation approach produces simulation domains suitable for flow visualization and allows for accurate characterization of petrophysical properties from non-destructive imaging data.

scholarly journals Comparative mapping of crawling-cell morphodynamics in deep learning-based feature space

2020 ◽  
Author(s):  
Daisuke Imoto ◽  
Nen Saito ◽  
Akihiko Nakajima ◽  
Gen Honda ◽  
Motohiko Ishida ◽  
...  
Keyword(s):  
Deep Learning ◽  
Feature Space ◽  
Real Data ◽  
Cell Polarization ◽  
Data Driven ◽  
Phase Field Modeling ◽  
Modeling Framework ◽  
Field Modeling ◽  
Quantitative Understanding ◽  
Low Dimensional

AbstractNavigation of fast migrating cells such as amoeba Dictyostelium and immune cells are tightly associated with their morphologies that range from steady polarized forms that support high directionality to those more complex and variable when making frequent turns. Model simulations are essential for quantitative understanding of these features and their origins, however systematic comparisons with real data are underdeveloped. Here, by employing deep-learning-based feature extraction combined with phase-field modeling framework, we show that a low dimensional feature space for 2D migrating cell morphologies obtained from the shape stereotype of keratocytes, Dictyostelium and neutrophils can be fully mapped by interlinked signaling network of cell-polarization and protrusion dynamics. Our analysis links the data-driven shape analysis to the underlying causalities by identifying key parameters critical for migratory morphologies both normal and aberrant under genetic and pharmacological perturbations. The results underscore the importance of deciphering self-organizing states and their interplay when characterizing morphological phenotypes.

Sign in / Sign up

Export Citation Format

Share Document