scholarly journals A Hybrid Data-Driven Deep Learning Technique for Fluid-Structure Interaction

2019 ◽  
Author(s):  
T. P. Miyanawala ◽  
Rajeev K. Jaiman
Keyword(s):  
Time Series ◽  
Deep Learning ◽  
Flow Field ◽  
Bluff Body ◽  
Dimensional Subspace ◽  
Data Driven ◽  
Long Time ◽  
Hybrid Data ◽  
Low Dimensional Approximations ◽  
Low Dimensional

Abstract This paper is concerned with the development of a hybrid data-driven technique for unsteady fluid-structure interaction systems. The proposed data-driven technique combines the deep learning framework with a projection-based low-order modeling. While the deep learning provides low-dimensional approximations from datasets arising from black-box solvers, the projection-based model constructs the low-dimensional approximations by projecting the original high-dimensional model onto a low-dimensional subspace. Of particular interest of this paper is to predict the long time series of unsteady flow fields of a freely vibrating bluff-body subjected to wake-body synchronization. We consider convolutional neural networks (CNN) for the learning dynamics of wake-body interaction, which assemble layers of linear convolutions with nonlinear activations to automatically extract the low-dimensional flow features. Using the high-fidelity time series data from the stabilized finite element Navier-Stokes solver, we first project the dataset to a low-dimensional subspace by proper orthogonal decomposition (POD) technique. The time-dependent coefficients of the POD subspace are mapped to the flow field via a CNN with nonlinear rectification, and the CNN is iteratively trained using the stochastic gradient descent method to predict the POD time coefficient when a new flow field is fed to it. The time-averaged flow field, the POD basis vectors, and the trained CNN are used to predict the long time series of the flow fields and the flow predictions are quantitatively assessed with the full-order (high-dimensional) simulation data. The proposed POD-CNN model based on the data-driven approximation has a remarkable accuracy in the entire fluid domain including the highly nonlinear near wake region behind a freely vibrating bluff body.

Bayesian recurrent neural network as a tool for reconstructing dynamical systems from multidimensional data

2020 ◽  
Author(s):  
Alexander Feigin ◽  
Aleksei Seleznev ◽  
Dmitry Mukhin ◽  
Andrey Gavrilov ◽  
Evgeny Loskutov
Keyword(s):  
Neural Network ◽  
Time Series ◽  
Cost Function ◽  
Recurrent Neural Network ◽  
Bayesian Optimization ◽  
Data Driven ◽  
High Dimensional ◽  
Model Learning ◽  
Low Dimensional ◽  
The Cost

<p>We suggest a new method for construction of data-driven dynamical models from observed multidimensional time series. The method is based on a recurrent neural network (RNN) with specific structure, which allows for the joint reconstruction of both a low-dimensional embedding for dynamical components in the data and an operator describing the low-dimensional evolution of the system. The key link of the method is a Bayesian optimization of both model structure and the hypothesis about the data generating law, which is needed for constructing the cost function for model learning.  The form of the model we propose allows us to construct a stochastic dynamical system of moderate dimension that copies dynamical properties of the original high-dimensional system. An advantage of the proposed method is the data-adaptive properties of the RNN model: it is based on the adjustable nonlinear elements and has easily scalable structure. The combination of the RNN with the Bayesian optimization procedure efficiently provides the model with statistically significant nonlinearity and dimension.<br>The method developed for the model optimization aims to detect the long-term connections between system’s states – the memory of the system: the cost-function used for model learning is constructed taking into account this factor. In particular, in the case of absence of interaction between the dynamical component and noise, the method provides unbiased reconstruction of the hidden deterministic system. In the opposite case when the noise has strong impact on the dynamics, the method yield a model in the form of a nonlinear stochastic map determining the Markovian process with memory. Bayesian approach used for selecting both the optimal model’s structure and the appropriate cost function allows to obtain the statistically significant inferences about the dynamical signal in data as well as its interaction with the noise components.<br>Data driven model derived from the relatively short time series of the QG3 model – the high dimensional nonlinear system producing chaotic behavior – is shown be able to serve as a good simulator for the QG3 LFV components. The statistically significant recurrent states of the QG3 model, i.e. the well-known teleconnections in NH, are all reproduced by the model obtained. Moreover, statistics of the residence times of the model near these states is very close to the corresponding statistics of the original QG3 model. These results demonstrate that the method can be useful in modeling the variability of the real atmosphere.</p><p>The work was supported by the Russian Science Foundation (Grant No. 19-42-04121).</p>

Reduced Order Model for Unsteady Fluid Flows via Recurrent Neural Networks

2019 ◽  
Author(s):  
Sandeep B. Reddy ◽  
Allan Ross Magee ◽  
Rajeev K. Jaiman ◽  
J. Liu ◽  
W. Xu ◽  
...  
Keyword(s):  
Neural Networks ◽  
Reynolds Number ◽  
Flow Field ◽  
Unsteady Flow ◽  
Recurrent Neural Networks ◽  
Reduced Order Model ◽  
Data Driven ◽  
Order Model ◽  
Reduced Order ◽  
Low Dimensional

Abstract In this paper, we present a data-driven approach to construct a reduced-order model (ROM) for the unsteady flow field and fluid-structure interaction. This proposed approach relies on (i) a projection of the high-dimensional data from the Navier-Stokes equations to a low-dimensional subspace using the proper orthogonal decomposition (POD) and (ii) integration of the low-dimensional model with the recurrent neural networks. For the hybrid ROM formulation, we consider long short term memory networks with encoder-decoder architecture, which is a special variant of recurrent neural networks. The mathematical structure of recurrent neural networks embodies a non-linear state space form of the underlying dynamical behavior. This particular attribute of an RNN makes it suitable for non-linear unsteady flow problems. In the proposed hybrid RNN method, the spatial and temporal features of the unsteady flow system are captured separately. Time-invariant modes obtained by low-order projection embodies the spatial features of the flow field, while the temporal behavior of the corresponding modal coefficients is learned via recurrent neural networks. The effectiveness of the proposed method is first demonstrated on a canonical problem of flow past a cylinder at low Reynolds number. With regard to a practical marine/offshore engineering demonstration, we have applied and examined the reliability of the proposed data-driven framework for the predictions of vortex-induced vibrations of a flexible offshore riser at high Reynolds number.

scholarly journals Embedding reconstruction methodology for short time series – application to large El Niño events

2010 ◽  
Vol 17 (6) ◽  
pp. 753-764 ◽  
Author(s):  
H. F. Astudillo ◽  
F. A. Borotto ◽  
R. Abarca-del-Rio
Keyword(s):  
Time Series ◽  
El Niño ◽  
Time Delays ◽  
El Nino ◽  
Southern Oscillation ◽  
Dimensional Subspace ◽  
Reconstruction Method ◽  
Short Time Series ◽  
Long Time ◽  
Short Time

Abstract. We propose an alternative approach for the embedding space reconstruction method for short time series. An m-dimensional embedding space is reconstructed with a set of time delays including the relevant time scales characterizing the dynamical properties of the system. By using a maximal predictability criterion a d-dimensional subspace is selected with its associated set of time delays, in which a local nonlinear blind forecasting prediction performs the best reconstruction of a particular event of a time series. An locally unfolded d-dimensional embedding space is then obtained. The efficiency of the methodology, which is mathematically consistent with the fundamental definitions of the local nonlinear long time-scale predictability, was tested with a chaotic time series of the Lorenz system. When applied to the Southern Oscillation Index (SOI) (observational data associated with the El Niño-Southern Oscillation phenomena (ENSO)) an optimal set of embedding parameters exists, that allows constructing the main characteristics of the El Niño 1982–1983 and 1997–1998 events, directly from measurements up to 3 to 4 years in advance.

scholarly journals A Data-Driven Long Time-Series Electrical Line Trip Fault Prediction Method Using an Improved Stacked-Informer Network

Sensors ◽  
2021 ◽  
Vol 21 (13) ◽  
pp. 4466
Author(s):  
Li Guo ◽  
Runze Li ◽  
Bin Jiang
Keyword(s):  
Time Series ◽  
Prediction Accuracy ◽  
Time Series Data ◽  
Short Term Memory ◽  
Fault Prediction ◽  
Data Driven ◽  
Series Data ◽  
Sequence Prediction ◽  
Long Time Series ◽  
Long Time

The monitoring of electrical equipment and power grid systems is very essential and important for power transmission and distribution. It has great significances for predicting faults based on monitoring a long sequence in advance, so as to ensure the safe operation of the power system. Many studies such as recurrent neural network (RNN) and long short-term memory (LSTM) network have shown an outstanding ability in increasing the prediction accuracy. However, there still exist some limitations preventing those methods from predicting long time-series sequences in real-world applications. To address these issues, a data-driven method using an improved stacked-Informer network is proposed, and it is used for electrical line trip faults sequence prediction in this paper. This method constructs a stacked-Informer network to extract underlying features of long sequence time-series data well, and combines the gradient centralized (GC) technology with the optimizer to replace the previously used Adam optimizer in the original Informer network. It has a superior generalization ability and faster training efficiency. Data sequences used for the experimental validation are collected from the wind and solar hybrid substation located in Zhangjiakou city, China. The experimental results and concrete analysis prove that the presented method can improve fault sequence prediction accuracy and achieve fast training in real scenarios.

scholarly journals A Case Study on Hospital Readmission Prediction Using Deep Learning Algorithms on EHRs

2021 ◽  
Vol 12 (3) ◽  
pp. 4267-4273
Author(s):  
B. Sushrith Et.al
Keyword(s):  
Time Series ◽  
Deep Learning ◽  
Financial Risk ◽  
Time Series Data ◽  
Learning Algorithms ◽  
Series Data ◽  
Critical Areas ◽  
Avoidable Readmission ◽  
Low Dimensional ◽  
Medical Concepts

In this paper, focus is made on predicting the patients who are going to be re-admitted back in the hospital before discharge using latest deep-learning algorithms is applied on the electronic health records of patients which is a time-series data. To begin with the study of the data and its analysis this project deployed the conventional supervised ML algorithms like the Logistic Regression, Naïve Bayes, Random Forest and SVM and compared their performances on different portion sizes of dataset. The final model built uses deep-learning architectures such as RNN and LSTM to improve the prediction results taking advantage of the time series data. Another feature added has been of low dimensional descriptions of medical concepts as the input to the model. Ultimately, this work tests, validates, and explains the developed system using the MIMIC-III dataset, which contains around 38000 patient’s information and about 61,155 patient’s data who admitted in ICU, duration of 10 years. The support from this exhaustive dataset is used to train the models that provide healthcare workers with proper information regarding their discharge and readmission in hospitals. These ML and deep learning models are used to know about the patient who is getting to be readmitted in the ICU before his discharge will help the hospital to allocate resources properly and also reduce the financial risk of patients. In order to reduce ICU readmission that can be avoided, hospitals have to be able to recognize patients who have a higher risk of ICU readmission. Those patients can then continue to stay in the ICU so that they will not have the risk of getting admit back to the hospital. Also, the resources of hospitals that were required for avoidable readmission can be re-allocated to more critical areas in the hospital that need them. A more effective model of predicting readmission system can play an important role in helping hospitals and ICU doctors to find the patients who are going to be readmitted before discharge. To build this system here we use different ML and deep-learning algorithms. Predictive models based on huge amounts of data are made to predict the patients who are going to be admitted back in the hospital after discharge.

scholarly journals Contributions towards Data driven Deep Learning methods to predict Steady State Fluid Flow in mechanical Heart Valves

2021 ◽  
Vol 7 (2) ◽  
pp. 625-628
Author(s):  
Jan Oldenburg ◽  
Julian Renkewitz ◽  
Michael Stiehm ◽  
Klaus-Peter Schmitz
Keyword(s):  
Neural Network ◽  
Deep Learning ◽  
Steady State ◽  
Flow Field ◽  
Heart Valves ◽  
Data Driven ◽  
High Shear ◽  
Blood Flows ◽  
The Neural Network ◽  
Data Driven Approach

Abstract It is commonly accepted that hemodynamic situation is related with cardiovascular diseases as well as clinical post-procedural outcome. In particular, aortic valve stenosis and insufficiency are associated with high shear flow and increased pressure loss. Furthermore, regurgitation, high shear stress and regions of stagnant blood flow are presumed to have an impact on clinical result. Therefore, flow field assessment to characterize the hemodynamic situation is necessary for device evaluation and further design optimization. In-vitro as well as in-silico fluid mechanics methods can be used to investigate the flow through prostheses. In-silico solutions are based on mathematical equitation’s which need to be solved numerically (Computational Fluid Dynamics - CFD). Fundamentally, the flow is physically described by Navier-Stokes. CFD often requires high computational cost resulting in long computation time. Techniques based on deep-learning are under research to overcome this problem. In this study, we applied a deep-learning strategy to estimate fluid flows during peak systolic steady-state blood flows through mechanical aortic valves with varying opening angles in randomly generated aortic root geometries. We used a data driven approach by running 3,500 two dimensional simulations (CFD). The simulation data serves as training data in a supervised deep learning framework based on convolutional neural networks analogous to the U-net architecture. We were able to successfully train the neural network using the supervised data driven approach. The results showing that it is feasible to use a neural network to estimate physiological flow fields in the vicinity of prosthetic heart valves (Validation error below 0.06), by only giving geometry data (Image) into the Network. The neural network generates flow field prediction in real time, which is more than 2500 times faster compared to CFD simulation. Accordingly, there is tremendous potential in the use of AIbased approaches predicting blood flows through heart valves on the basis of geometry data, especially in applications where fast fluid mechanic predictions are desired.

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.

scholarly journals TapNet: Multivariate Time Series Classification with Attentional Prototypical Network

2020 ◽  
Vol 34 (04) ◽  
pp. 6845-6852 ◽  
Author(s):  
Xuchao Zhang ◽  
Yifeng Gao ◽  
Jessica Lin ◽  
Chang-Tien Lu
Keyword(s):  
Time Series ◽  
Deep Learning ◽  
Time Series Data ◽  
Multivariate Time Series ◽  
Feature Representation ◽  
Series Data ◽  
Time Series Classification ◽  
Random Group ◽  
Low Dimensional ◽  
Low Dimensional Features

With the advance of sensor technologies, the Multivariate Time Series classification (MTSC) problem, perhaps one of the most essential problems in the time series data mining domain, has continuously received a significant amount of attention in recent decades. Traditional time series classification approaches based on Bag-of-Patterns or Time Series Shapelet have difficulty dealing with the huge amounts of feature candidates generated in high-dimensional multivariate data but have promising performance even when the training set is small. In contrast, deep learning based methods can learn low-dimensional features efficiently but suffer from a shortage of labelled data. In this paper, we propose a novel MTSC model with an attentional prototype network to take the strengths of both traditional and deep learning based approaches. Specifically, we design a random group permutation method combined with multi-layer convolutional networks to learn the low-dimensional features from multivariate time series data. To handle the issue of limited training labels, we propose a novel attentional prototype network to train the feature representation based on their distance to class prototypes with inadequate data labels. In addition, we extend our model into its semi-supervised setting by utilizing the unlabeled data. Extensive experiments on 18 datasets in a public UEA Multivariate time series archive with eight state-of-the-art baseline methods exhibit the effectiveness of the proposed model.

Sign in / Sign up

Export Citation Format

Share Document