Skip to main content

INVERSION EN TIEMPO REAL CON METODOS DE APRENDIZAJE PROFUNDO

PID2019-108111RB-I00

•
Funding agency name Agencia Estatal de Investigación
Funding agency acronym AEI
Program Programa Estatal de I+D+i Orientado a los Retos de la Sociedad
Call Proyectos de I+D+i (Generación de Conocimiento y Retos Investigación)
Call year 2019
State Plan 2017-2020 Plan Estatal de Investigación Científica y Técnica y de Innovación
Beneficiary centre ASOC BCAM - BASQUE CENTER FOR APPLIED MATHEMATICS

Resultados relacionados

Resultados totales (Incluyendo duplicados): 52
Encontrada(s) 1 página(s)

BIRD. BCAM's Institutional Repository Data
Publication Artículo científico (JournalArticle)

Bearing assessment tool for longitudinal bridge performance

  • Garcia-Sanchez, D.
  • Fernandez-Navamuel, A.
  • Zamora, D.
  • Alvear, D.
  • Pardo, D.
This work provides an unsupervised learning approach based on a single-valued performance indicator to monitor the global behavior of critical components in a viaduct, such as bearings. We propose an outlier detection method for longitudinal displacements to assess the behavior of a singular asymmetric prestressed concrete structure with a 120 m high central pier acting as a fixed point. We first show that the available long-term horizontal displacement measurements recorded during the undamaged state exhibit strong correlations at the different locations of the bearings. Thus, we combine measurements from four sensors to design a robust performance indicator that is only weakly affected by temperature variations after the application of Principal Component Analysis. We validate the method and show its efficiency against false positives and negatives by using several metrics: accuracy, precision, recall and F1 score. Due to its unsupervised learning scope, the proposed technique is intended to serve as a real-time supervision tool that complements maintenance inspections. It aims to provide support for the prioritization and postponement of maintenance actions in bridge management.
BIRD. BCAM's Institutional Repository Data
Publication Artículo científico (JournalArticle)

Sensitivity and Uncertainty Analysis by Discontinuous Galerkin of Lock-in Thermography for Crack Characterization

  • Omella, A. J.
  • Celorrio, R.
  • Pardo, D.
This work focuses on the characterization of narrow vertical cracks of nite size using optically excited lock-in thermography (OLT). To characterize these cracks, we need to solve an ill-posed inverse problem. As a previous step to the solution of this inverse problem, we propose a sensitivity analysis to quantify the influence that the parameters involved in the model have on the surface temperature. Some of these parameters are estimated at the laboratory and they incorporate uncertainty that may severely aff ect the reconstruction of thin cracks. For this reason, we design a calibration criterion based on the sensitivity analysis to determine which parameters we need to include as unknowns of the inverse problem. We perform this analysis using a numerical discontinuous Galerkin method. Additionally, we propose a theoretical noise model for the thermograms. Then, we use a weighted least square method (WLS) to determine the parameters from the experimental thermograms. We also obtain a theoretical uncertainty of the reconstructed parameters in OLT-WLS fi tting according to the used surface temperature dataset. Finally, we perform a numerical experiment with a 2.4 m-thick vertical crack to show the sensitivities of the surface temperature with respect to the model parameters. We also determine the uncertainty of the parameters under di fferent datasets with known noise characteristics.
BIRD. BCAM's Institutional Repository Data
Publication Artículo científico (JournalArticle)

Rational approximation of P-wave kinematics — Part 1: Transversely isotropic media

  • Abedi, M.M.
In seismic data processing and several wave propagation modeling algorithms, the phase velocity, group velocity, and traveltime equations are essential. To have these equations in explicit form, or to reduce algebraic complexity, approximation methods are used. For the approximation of P-wave kinematics in acoustic transversely isotropic media, we have developed a new flexible 2D functional equation in a continued fraction form. Using different orders of the continued fraction, we obtain different approximations for (1) phase velocity as a function of phase direction, (2) group velocity as a function of group direction, and (3) traveltime as a function of offset. Then, we use them in the approximation of the group direction as a function of phase direction, and phase direction as a function of group direction. The proposed approximations have a rational form, which is considered algebraically simple and computationally efficient. The used continued fraction form rapidly converges to exact kinematics. By introducing the optimal ray into our approximations and using it for parameter definition, the convergence becomes faster, so the accuracy of the existing most accurate approximations is available by the third order, and new most accurate approximations are obtained by the fourth order of the proposed general form. The error of the most accurate version of the proposed approximations is below 0.001% for moderate anisotropic models with an anellipticity parameter up to 0.3. This high accuracy is considered to be attractive in practical implementations that use the kinematic equations and their derivatives., European Regional Development Fund (ERDF) through the Interreg V-A Spain-France-Andorra program POCTEFA 2014-2020 Project PIXIL (EFA362/19). The grant "Artificial Intelligence in BCAM number EXP. 2019/00432"
BIRD. BCAM's Institutional Repository Data
Publication Artículo científico (JournalArticle)

Equivalence between the DPG method and the Exponential Integrators for linear parabolic problems

  • Muñoz-Matute, J.
  • Pardo, D.
  • Demkowicz, L.
The Discontinuous Petrov-Galerkin (DPG) method and the exponential integrators are two well established numerical methods for solving Partial Di fferential Equations (PDEs) and sti ff systems of Ordinary Di fferential Equations (ODEs), respectively. In this work, weapply the DPG method in the time variable for linear parabolic problems and we calculate the optimal test functions analytically. We show that the DPG method in time is equivalent to exponential integrators for the trace variables, which are decoupled from the interior variables. In addition, the DPG optimal test functions allow us to compute the approximated solutions in the time element interiors. This DPG method in time allows to construct a posteriori error estimations in order to perform adaptivity. We generalize this novel DPG-based time-marching scheme to general fi rst order linear systems of ODEs. We show the performance of the proposed method for 1D and 2D + time linear parabolic PDEs after discretizing in space by the nite element method.
BIRD. BCAM's Institutional Repository Data
Publication Artículo científico (JournalArticle)

A new acoustic assumption for orthorhombic media

  • Abedi, M.M.
  • Stovas, A.
In exploration seismology, the acquisition, processing and inversion of P-wave data is a routine. However, in orthorhombic anisotropic media, the governing equations that describe the P-wave propagation are coupled with two S waves that are considered as redundant noise. The main approach to free the P-wave signal from the S-wave noise is the acoustic assumption on the wave propagation. The conventional acoustic assumption for orthorhombic media zeros out the S-wave velocities along three orthogonal axes, but leaves significant S-wave artefacts in all other directions. The new acoustic assumption that we propose mitigates the S-wave artefacts by zeroing out their velocities along the three orthogonal symmetry planes of orthorhombic media. Similar to the conventional approach, our method reduces the number of required model parameters from nine to six. As numerical experiments on multiple orthorhombic models show, the accuracy of the new acoustic assumption also compares well to the conventional approach. On the other hand, while the conventional acoustic assumption simplifies the governing equations, the new acoustic assumption further complicates them—an issue that emphasizes the necessity of simple approximate equations. Accordingly, we also propose simpler rational approximate phase-velocity and eikonal equations for the new acoustic orthorhombic media. We show a simple ray tracing example and find out that the proposed approximate equations are still highly accurate., Eusko Jaularitza -- Basque Government : EXP. 2019/00432
BIRD. BCAM's Institutional Repository Data
Publication Artículo científico (JournalArticle)

Rational approximation of P-wave kinematics — Part 2: Orhorhombic media

  • Abedi, M.M.
Orthorhombic anisotropy is a modern standard for 3D seismic studies in complex geologic settings. Several seismic data processing methods and wave propagation modeling algorithms in orthorhombic media rely on phase-velocity, group-velocity, and traveltime approximations. The algebraic simplicity of an approximate equation is an important factor in these media because the governing equations are more complicated than transversely isotropic media. To approximate the P-wave kinematics in acoustic orthorhombic media, we have developed a new 3D general functional equation that has a simple rational form. Using the general form, we adopt two versions of rational approximations for the phase velocity, group velocity, and traveltime. The first version uses a simpler functional form and parameter definition within the orthorhombic symmetry planes. The second version is more accurate, using one parameter that is defined out of the symmetry planes. For the phase velocity, we obtain another approximation that is no longer rational but is still algebraically simple, exact for 3D transversely isotropic media, and it is exact within the symmetry planes of orthorhombic media. We find superior accuracy in our approximations compared with previous ones, using numerical studies on multiple moderately anisotropic orthorhombic models. We investigate the effect of the negative anellipticity parameters on the accuracy and find that, in models in which the error of the existing most accurate approximations exceeds 2%, the error of the new approximations remains below 0.2%. The adopted approximations are algebraically simpler and stably more accurate than existing approximations; therefore, they may be considered as attractive alternatives for the existing approximations in many practical applications. We extend the applicability of our approximations by using them to obtain the equations of group direction as a function of phase direction and vice versa, which are useful in wave propagation modeling methods., the grant "Artificial Intelligence in BCAM number EXP. 2019/00432", the European Regional Development Fund (ERDF) through the Interreg V-A Spain-France-Andorra program POCTEFA 2014-2020 Project PIXIL (EFA362/19);
BIRD. BCAM's Institutional Repository Data
Publication Artículo científico (JournalArticle)

Goal-oriented adaptivity for a conforming residual minimization method in a dual discontinuous Galerkin norm

  • Rojas, S.
  • Pardo, D.
  • Behnoudfare, P.
  • Calo, V. M.
We propose a goal-oriented mesh-adaptive algorithm for a finite element method stabilized via residual minimization on dual discontinuous-Galerkin norms. By solving a saddle-point problem, this residual minimization delivers a stable continuous approximation to the solution on each mesh instance and a residual projection onto a broken polynomial space, which is a robust error estimator to minimize the discrete energy norm via automatic mesh refinement. In this work, we propose and analyze a goal-oriented adaptive algorithm for this stable residual minimization. We solve the primal and adjoint problems considering the same saddle-point formulation and different right-hand sides. By solving a third stable problem, we obtain two efficient error estimates to guide goal oriented adaptivity. We illustrate the performance of this goal-oriented adaptive strategy on advection-diffusion reaction problems.
BIRD. BCAM's Institutional Repository Data
Publication Artículo científico (JournalArticle)

Refined isogeometric analysis for generalized Hermitian eigenproblems

  • Hashemian, A.
  • Pardo, D.
  • Calo, V.M.
We use refined isogeometric analysis (rIGA) to solve generalized Hermitian eigenproblems (Ku = λMu). rIGA conserves the desirable properties of maximum-continuity isogeometric analysis (IGA) while it reduces the solution cost by adding zero-continuity basis functions, which decrease the matrix connectivity. As a result, rIGA enriches the approximation space and reduces the interconnection between degrees of freedom. We compare computational costs of rIGA versus those of IGA when employing a Lanczos eigensolver with a shift-and-invert spectral transformation. When all eigenpairs within a given interval [λ_s,λ_e] are of interest, we select several shifts σ_k ∈ [λ_s,λ_e] using a spectrum slicing technique. For each shift σ_k, the factorization cost of the spectral transformation matrix K − σ_k M controls the total computational cost of the eigensolution. Several multiplications of the operator matrix (K − σ_k M)^−1 M by vectors follow this factorization. Let p be the polynomial degree of the basis functions and assume that IGA has maximum continuity of p−1. When using rIGA, we introduce C^0 separators at certain element interfaces to minimize the factorization cost. For this setup, our theoretical estimates predict computational savings to compute a fixed number of eigenpairs of up to O(p^2) in the asymptotic regime, that is, large problem sizes. Yet, our numerical tests show that for moderate-size eigenproblems, the total observed computational cost reduction is O(p). In addition, rIGA improves the accuracy of every eigenpair of the first N_0 eigenvalues and eigenfunctions, where N_0 is the total number of modes of the original maximum-continuity IGA discretization.
BIRD. BCAM's Institutional Repository Data
Publication Artículo científico (JournalArticle)

Massive Database Generation for 2.5D Borehole Electromagnetic Measurements using Refined Isogeometric Analysis

  • Hashemian, A.
  • Garcia, D.
  • Rivera, J.A.
  • Pardo, D.
Borehole resistivity measurements are routinely inverted in real-time during geosteering operations. The inversion process can be efficiently performed with the help of advanced artificial intelligence algorithms such as deep learning. These methods require a massive dataset that relates multiple Earth models with the corresponding borehole resistivity measurements. In here, we propose to use an advanced numerical method —refined isogeometric analysis (rIGA)— to perform rapid and accurate 2.5D simulations and generate databases when considering arbitrary 2D Earth models. Numerical results show that we can generate a meaningful synthetic database composed of 100,000 Earth models with the corresponding measurements in 56 hours using a workstation equipped with two CPUs., European POCTEFA 2014–2020 Project PIXIL (EFA362/19); The grant ‘‘Artificial Intelligence in BCAM number EXP. 2019/00432
BIRD. BCAM's Institutional Repository Data
Publication Artículo científico (JournalArticle)

Variance Reduction with Array-RQMC for Tau-Leaping Simulation of Stochastic Biological and Chemical Reaction Networks

  • Puchhammer, F.
  • Ben Abdellah, A.
  • L'Ecuyer, P.
We explore the use of Array-RQMC, a randomized quasi-Monte Carlo method designed for the simulation of Markov chains, to reduce the variance when simulating stochastic biological or chemical reaction networks with $\tau$-leaping. The task is to estimate the expectation of a function of molecule copy numbers at a given future time $T$ by the sample average over $n$ sample paths, and the goal is to reduce the variance of this sample-average estimator. We find that when the method is properly applied, variance reductions by factors in the thousands can be obtained. These factors are much larger than those observed previously by other authors who tried RQMC methods for the same examples. Array-RQMC simulates an array of realizations of the Markov chain and requires a sorting function to reorder these chains according to their states, after each step. The choice of sorting function is a key ingredient for the efficiency of the method, although in our experiments, Array-RQMC was never worse than ordinary Monte Carlo, regardless of the sorting method. The expected number of reactions of each type per step also has an impact on the efficiency gain., ERDF, ESF, EXP. 2019/00432, Canada Research Chair, IVADO Research Grant, NSERC Discovery Grant RGPIN-110050.
BIRD. BCAM's Institutional Repository Data
Publication Artículo científico (JournalArticle)

Monte Carlo and Quasi-Monte Carlo Density Estimation via Conditioning

  • L'Ecuyer, P.
  • Puchhammer, F.
  • Ben Abdellah, A.
Estimating the unknown density from which a given independent sample originates is more difficult than estimating the mean, in the sense that for the best popular non-parametric density estimators, the mean integrated square error converges more slowly than at the canonical rate of $\mathcal{O}(1/n)$. When the sample is generated from a simulation model and we have control over how this is done, we can do better. We examine an approach in which conditional Monte Carlo yields, under certain conditions, a random conditional density which is an unbiased estimator of the true density at any point. By averaging independent replications, we obtain a density estimator that converges at a faster rate than the usual ones. Moreover, combining this new type of estimator with randomized quasi-Monte Carlo to generate the samples typically brings a larger improvement on the error and convergence rate than for the usual estimators, because the new estimator is smoother as a function of the underlying uniform random numbers., IVADO Research Grant, NSERC-Canada Discorvery Grant, Canada Research Chair, Inria International Chair, ERDF, ESF, EXP. 2019/00432
BIRD. BCAM's Institutional Repository Data
Publication Contribución de congreso (ConferenceOutput)

Density Estimation by Monte Carlo and Quasi-Monte Carlo

  • L'Ecuyer, P.
  • Puchhammer, F.
Estimating the density of a continuous random variable $X$ has been studied extensively in statistics, in the setting where $n$ independent observations of $X$ are given a priori and one wishes to estimate the density from that. Popular methods include histograms and kernel density estimators. In this review paper, we are interested instead in the situation where the observations are generated by Monte Carlo simulation from a model. Then, one can take advantage of variance reduction methods such as stratification, conditional Monte Carlo, and randomized quasi-Monte Carlo (RQMC), and obtain a more accurate density estimator than with standard Monte Carlo for a given computing budget. We discuss several ways of doing this, proposed in recent papers, with a focus on methods that exploit RQMC. A first idea is to directly combine RQMC with a standard kernel density estimator. Another one is to adapt a simulation-based derivative estimation method such as smoothed perturbation analysis or the likelihood ratio method to obtain a continuous estimator of the cumulative density function (CDF), whose derivative is an unbiased estimator of the density. This can then be combined with RQMC. We summarize recent theoretical results with these approaches and give numerical illustrations of how they improve the convergence of the mean square integrated error., NSERC Discovery Grant, IVADO Grant, ERDF, ESF, EXP. 2019/00432
BIRD. BCAM's Institutional Repository Data
Publication Contribución de congreso (ConferenceOutput)

A Tool for Custom Construction of QMC and RQMC Point Sets

  • Godin, M.
  • Marion, P.
  • L'Ecuyer, P.
  • Puchhammer, F.
We present LatNet Builder, a software tool to find good parameters for lattice rules, polynomial lattice rules, and digital nets in base 2, for quasi-Monte Carlo (QMC) and randomized quasi-Monte Carlo (RQMC) sampling over the s-dimensional unit hypercube. The selection criteria are figures of merit that give different weights to different subsets of coordinates. They are upper bounds on the worst-case error (for QMC) or variance (for RQMC) for integrands rescaled to have a norm of at most one in certain Hilbert spaces of functions. We summarize what are the various Hilbert spaces, discrepancies, types of weights, figures of merit, types of constructions, and search methods supported by LatNet Builder. We briefly discuss its organization and we provide simple illustrations of what it can do., NSERC Discovery Grant, IVADO Grant, Corps des Mines Stipend, ERDF, ESF, EXP. 2019/00432
BIRD. BCAM's Institutional Repository Data
Publication Capítulo o parte de un libro (BookPart)

Multiscale Modelling and Simulation of Advanced Battery Materials

  • Rincón, M.
  • García-Daza, F.
  • Fernández-Pendás, M.
  • Carrasco, J.
  • Akhmatskaya, E.
Development of efficient strategies for the rational design of materials involved in the production and storage of renewable energy is essential for accelerating the transition to a low-carbon economy. To contribute to this goal, we propose a novel workflow for the assessment and optimization of battery materials. The approach effectively combines quantum and atomistic modelling/simulations, enhanced by efficient sampling, Bayesian parameterization, and experimental information. It is implemented to study prospective materials for lithium and sodium batteries.
BIRD. BCAM's Institutional Repository Data
Publication Artículo científico (JournalArticle)

Approximations for traveltime, slope, curvature, and geometrical spreading of elastic waves in layered transversely isotropic media

  • Abedi, M.M.
  • Pardo, D.
  • Stovas, A.
Each seismic body wave, including quasi compressional, shear, and converted wave modes, carries useful subsurface information. For processing, imaging, amplitude analysis, and forward modeling of each wave mode, we need approximate equations of traveltime, slope (ray-parameter), and curvature as a function of offset. Considering the large offset coverage of modern seismic acquisitions, we propose new approximations designed to be accurate at zero and infinitely large offsets over layered transversely isotropic media with vertical symmetry axis (VTI). The proposed approximation for traveltime is a modified version of the extended generalized moveout approximation that comprises six parameters. The proposed direct approximations for ray-parameter and curvature use new, algebraically simple, equations with three parameters. We define these parameters for each wave mode without ray tracing so that we have similar approximate equations for all wave modes that only change based on the parameter definitions. However, our approximations are unable to reproduce S-wave triplications that may occur in some strongly anisotropic models. Using our direct approximation of traveltime derivatives, we also obtain a new expression for the relative geometrical spreading. We demonstrate the high accuracy of our approximations using numerical tests on a set of randomly generated multilayer models. Using synthetic data, we present simple applications of our approximations for normal moveout correction and relative geometrical spreading compensation of different wave modes., POCTEFA 2014-2020 Project PIXIL (EFA362/19)
BIRD. BCAM's Institutional Repository Data
Publication Artículo científico (JournalArticle)

A Finite Element based Deep Learning solver for parametric PDEs

  • Uriarte, C.
  • Pardo, D.
  • Omella, A. J.
We introduce a dynamic Deep Learning (DL) architecture based on the Finite Element Method (FEM) to solve linear parametric Partial Differential Equations(PDEs). The connections between neurons in the architecture mimic the Finite Element connectivity graph when applying mesh refinements. We select and discuss several losses employing preconditioners and different norms to enhance convergence. For simplicity, we implement the resulting Deep-FEM in one spatial domain (1D), although its extension to 2D and 3D problems is straightforward. Extensive numerical experiments show in general good approximations for both symmetric positive definite (SPD) and indefinite problems in parametric and non-parametric problems. However, in some cases, lack of convexity prevents us from obtaining high-accuracy solutions., This work has received funding from: the European Union's Horizon 2020 research and innovation program under the Marie Sklodowska-Curie grant agreement No 777778 (MATHROCKS); the European Regional Development Fund (ERDF) through the Interreg V-A Spain-France-Andorra program POCTEFA 2014-2020 Project PIXIL (EFA362/19); the Spanish Ministry of Science and Innovation projects with references PID2019-108111RB-I00 (FEDER/AEI) and PDC2021-121093-I00, the "BCAM Severo Ochoa" accreditation of excellence (SEV-2017-0718); and the Basque Government through the BERC 2018-2021 program, the three Elkartek projects 3KIA (KK-2020/00049), EXPERTIA (KK-2021/00048), and SIGZE (KK-2021/00095), and the Consolidated Research Group MATHMODE (IT1294-19) given by the Department of Education.
BIRD. BCAM's Institutional Repository Data
Publication Artículo científico (JournalArticle)

Error representation of the time-marching DPG scheme

  • Muñoz-Matute, J.
  • Demkowicz, Leszek
  • Pardo, D.
In this article, we introduce an error representation function to perform adaptivity in time of the recently developed time-marching Discontinuous Petrov–Galerkin (DPG) scheme. We first provide an analytical expression for the error that is the Riesz representation of the residual. Then, we approximate the error by enriching the test space in such a way that it contains the optimal test functions. The local error contributions can be efficiently computed by adding a few equations to the time-marching scheme. We analyze the quality of such approximation by constructing a Fortin operator and providing an a posteriori error estimate. The time-marching scheme proposed in this article provides an optimal solution along with a set of efficient and reliable local error contributions to perform adaptivity. We validate our method for both parabolic and hyperbolic problems.
BIRD. BCAM's Institutional Repository Data
Publication Artículo científico (JournalArticle)

Deep learning enhanced principal component analysis for structural health monitoring

  • Fernandez-Navamuel, A.
  • Magalhães, Filipe
  • Zamora-Sánchez, Diego
  • Omella, Ángel J.
  • Garcia-Sanchez, David
  • Pardo, D.
This paper proposes a Deep Learning Enhanced Principal Component Analysis (PCA) approach for outlier detection to assess the structural condition of bridges. We employ partially explainable autoencoder architecture to replicate and enhance the data compression and reconstruction ability of PCA. The particularity of the method lies in the addition of residual connections to account for nonlinearities. We apply the proposed method to monitoring data obtained from two bridges under real operation conditions and compare the results before and after adding the residual connections. Results show that the addition of residual connections enhances the outlier detection ability of the network, allowing to detect lighter damages.
BIRD. BCAM's Institutional Repository Data
Publication Artículo científico (JournalArticle)

On quadrature rules for solving Partial Differential Equations using Neural Networks

  • Rivera, J.A
  • Taylor, J.M.
  • Omella, Ángel J.
  • Pardo, D.
Neural Networks have been widely used to solve Partial Differential Equations. These methods require to approximate definite integrals using quadrature rules. Here, we illustrate via 1D numerical examples the quadrature problems that may arise in these applications and propose several alternatives to overcome them, namely: Monte Carlo methods, adaptive integration, polynomial approximations of the Neural Network output, and the inclusion of regularization terms in the loss. We also discuss the advantages and limitations of each proposed numerical integration scheme. We advocate the use of Monte Carlo methods for high dimensions (above 3 or 4), and adaptive integration or polynomial approximations for low dimensions (3 or below). The use of regularization terms is a mathematically elegant alternative that is valid for any spatial dimension; however, it requires certain regularity assumptions on the solution and complex mathematical analysis when dealing with sophisticated Neural Networks.
BIRD. BCAM's Institutional Repository Data
Publication Artículo científico (JournalArticle)

Supervised Deep Learning with Finite Element simulations for damage identification in bridges

  • Fernandez-Navamuel, A.
  • Zamora-Sánchez, Diego
  • Omella, Ángel J.
  • Pardo, D.
  • Garcia-Sanchez, David
  • Magalhães, Filipe
This work proposes a supervised Deep Learning approach for damage identification in bridge structures. We employ a hybrid methodology that incorporates Finite Element simulations to enrich the training phase of a Deep Neural Network with synthetic damage scenarios. The neural network is based on autoencoders and its particular architecture allows to activate or deactivate nonlinear connections under need. The methodology intends to contribute to the progress towards the applicability of Structural Health Monitoring practices in full-scale bridge structures. The ultimate goal is to estimate the location and severity of damage from measurements of the dynamic response of the structure. The damages we seek to detect correspond to material degradations that affect wide areas of the structure by reducing its stiffness properties. Our method allows a feasible adaptation to large systems with complex parametrizations and structural particularities. We investigate the performance of the proposed method on two full-scale instrumented bridges, obtaining adequate results for the testing datasets even in presence of measurement uncertainty. Besides, the method successfully predicts the damage condition for two real damage scenarios of increasing severity available in one of the bridges.
BIRD. BCAM's Institutional Repository Data
Publication Artículo científico (JournalArticle)

Exploiting the Kronecker product structure of φ−functions in exponential integrators

  • Muñoz-Matute, J.
  • Pardo, D.
  • Calo, Victor M.
Exponential time integrators are well-established discretization methods for time semilinear systems of ordinary differential equations. These methods use (Formula presented.) functions, which are matrix functions related to the exponential. This work introduces an algorithm to speed up the computation of the (Formula presented.) function action over vectors for two-dimensional (2D) matrices expressed as a Kronecker sum. For that, we present an auxiliary exponential-related matrix function that we express using Kronecker products of one-dimensional matrices. We exploit state-of-the-art implementations of (Formula presented.) functions to compute this auxiliary function's action and then recover the original (Formula presented.) action by solving a Sylvester equation system. Our approach allows us to save memory and solve exponential integrators of 2D+time problems in a fraction of the time traditional methods need. We analyze the method's performance considering different linear operators and with the nonlinear 2D+time Allen–Cahn equation.
BIRD. BCAM's Institutional Repository Data
Publication Artículo científico (JournalArticle)

Nonhyperbolic normal moveout stretch correction with deep learning automation

  • Abedi, M.M.
  • Pardo, D.
Normal-moveout (NMO) correction is a fundamental step in seismic data processing. It consists of mapping seismic data from recorded traveltimes to corresponding zero-offset times. This process produces wavelet stretching as an undesired by-product. We have addressed the NMO stretching problem with two methods: (1) an exact stretch-free NMO correction that prevents the stretching of primary reflections and (2) an approximate post-NMO stretch correction. Our stretch-free NMO produces parallel moveout trajectories for primary reflections. Our post-NMO stretch correction calculates the moveout of stretched wavelets as a function of offset. Both methods are based on the generalized moveout approximation and are suitable for application in complex anisotropic or heterogeneous environments. We use new moveout equations, modify the original parameter functions to be a constant over the primary reflections, and then interpolate the seismogram amplitudes at the calculated traveltimes. For fast and automatic modification of the parameter functions, we use deep learning. We design a deep neural network (DNN) using convolutional layers and residual blocks. To train the DNN, we generate a set of 40,000 synthetic NMO-corrected common-midpoint gathers and the corresponding desired outputs of the DNN. The data set is generated using different velocity profiles, wavelets, and offset vectors, and it includes multiples, ground roll, and band-limited random noise. The simplicity of the DNN task - a 1D identification of primary reflections - improves the generalization in practice. We use the trained DNN and find successful applications of our stretch-correction method on synthetic and different real data sets.
BIRD. BCAM's Institutional Repository Data
Publication Artículo científico (JournalArticle)

Refined isogeometric analysis of quadratic eigenvalue problems

  • Hashemian, A.
  • Garcia, D.
  • Pardo, D.
  • Calo, V.M.
Certain applications that analyze damping effects require the solution of quadratic eigenvalue problems (QEPs). We use refined isogeometric analysis (rIGA) to solve quadratic eigenproblems. rIGA discretization, while conserving desirable properties of maximum-continuity isogeometric analysis (IGA), reduces the interconnection between degrees of freedom by adding low-continuity basis functions. This connectivity reduction in rIGA's algebraic system results in faster matrix LU factorizations when using multifrontal direct solvers. We compare computational costs of rIGA versus those of IGA when employing Krylov eigensolvers to solve quadratic eigenproblems arising in 2D vector-valued multifield problems. For large problem sizes, the eigencomputation cost is governed by the cost of LU factorization, followed by costs of several matrix–vector and vector–vector multiplications, which correspond to Krylov projections. We minimize the computational cost by introducing C0 and C1 separators at specific element interfaces for our rIGA generalizations of the curl-conforming Nédélec and divergence-conforming Raviart–Thomas finite elements. Let p be the polynomial degree of basis functions; the LU factorization is up to O((p−1)2) times faster when using rIGA compared to IGA in the asymptotic regime. Thus, rIGA theoretically improves the total eigencomputation cost by O((p−1)2) for sufficiently large problem sizes. Yet, in practical cases of moderate-size eigenproblems, the improvement rate deteriorates as the number of computed eigenvalues increases because of multiple matrix–vector and vector–vector operations. Our numerical tests show that rIGA accelerates the solution of quadratic eigensystems by O(p−1) for moderately sized problems when we seek to compute a reasonable number of eigenvalues.
BIRD. BCAM's Institutional Repository Data
Publication Artículo científico (JournalArticle)

Combining DPG in space with DPG time-marching scheme for the transient advection–reaction equation

  • Muñoz-Matute, J.
  • Demkowicz, L.
  • Roberts, N. V.
In this article, we present a general methodology to combine the Discontinuous Petrov-Galerkin (DPG) method in space and time in the context of methods of lines for transient advection-reaction problems. We  rst introduce a semidiscretization in space with a DPG method rede ning the ideas of optimal testing and practicality of the method in this context. Then, we apply the recently developed DPG-based time-marching scheme, which is of exponential-type, to the resulting system of Ordinary Differential Equations (ODEs). We also discuss how to e ciently compute the action of the exponential of the matrix coming from the space semidiscretization without assembling the full matrix. Finally, we verify the proposed method for 1D+time advection-reaction problems showing optimal convergence rates for smooth solutions and more stable results for linear conservation laws comparing to the classical exponential integrators.
BIRD. BCAM's Institutional Repository Data
Publication Artículo científico (JournalArticle)

A Deep Double Ritz Method (D2RM) for solving Partial Differential Equations using Neural Networks

  • Uriarte, C.
  • Pardo, D.
  • Muga, I.
  • Muñoz-Matute, J.
Residual minimization is a widely used technique for solving Partial Differential Equations in variational form. It minimizes the dual norm of the residual, which naturally yields a saddle-point (min–max) problem over the so-called trial and test spaces. In the context of neural networks, we can address this min–max approach by employing one network to seek the trial minimum, while another network seeks the test maximizers. However, the resulting method is numerically unstable as we approach the trial solution. To overcome this, we reformulate the residual minimization as an equivalent minimization of a Ritz functional fed by optimal test functions computed from another Ritz functional minimization. We call the resulting scheme the Deep Double Ritz Method (DRM), which combines two neural networks for approximating trial functions and optimal test functions along a nested double Ritz minimization strategy. Numerical results on different diffusion and convection problems support the robustness of our method, up to the approximation properties of the networks and the training capacity of the optimizers.
BIRD. BCAM's Institutional Repository Data
Publication Artículo científico (JournalArticle)

A painless multi-level automatic goal-oriented hp-adaptive coarsening strategy for elliptic and non-elliptic problems

  • Caro, F.V.
  • Darrigrand, V.
  • Alvarez-Aramberri, J.
  • Alberdi, E.
  • Pardo, D.
This work extends an automatic energy-norm $hp$-adaptive strategy based on performing quasi-optimal unrefinements to the case of non-elliptic problems and goal-oriented adaptivity. The proposed approach employs a multi-level hierarchical data structure and alternates global $h$- and $p$-refinements with a coarsening step. Thus, at each unrefinement step, we eliminate the basis functions with the lowest contributions to the solution. When solving elliptic problems using energy-norm adaptivity, the removed basis functions are those with the lowest contributions to the energy of the solution. For non-elliptic problems or goal-oriented adaptivity, we propose an upper bound of the error representation expressed in terms of an inner product of the specific equation, leading to error indicators that deliver quasi-optimal $hp$-unrefinements. This unrefinement strategy removes unneeded unknowns possibly introduced during the pre-asymptotical regime. In addition, the grids over which we perform the unrefinements are arbitrary, and thus, we can limit their size and associated computational costs. We numerically analyze our algorithm for energy-norm and goal-oriented adaptivity. In particular, we solve two-dimensional ($2$D) Poisson, Helmholtz, convection-dominated equations, and a three-dimensional ($3$D) Helmholtz-like problem. In all cases, we observe \revb{exponential} convergence rates. Our algorithm is robust and straightforward to implement; therefore, it can be easily adapted for industrial applications., BERC.2022-2025
BIRD. BCAM's Institutional Repository Data
Publication Artículo científico (JournalArticle)

A Multidirectional Deep Neural Network for Self-Supervised Reconstruction of Seismic Data

  • Abedi, M.M.
  • Pardo, D.
Seismic studies exhibit gaps in the recorded data due to surface obstacles. To fill in the gaps with self-supervised deep learning, the network learns to predict different events from the recorded parts of data and then applies it to reconstruct the missing parts of the same dataset. We propose two improvements to the task: a rearrangement of the data, and a new deep-learning approach. We rearrange the traces of a 2D acquisition line as 3D data cubes, sorting the traces by the source and receiver coordinates. This 3D representation offers more information about the structure of the seismic events and allows a coherent reconstruction of them. However, learning the structure of events in 3D cubes is more complicated than in 2D images while the size of the training dataset is limited. Thus, we propose a specific architecture and training strategy to take advantage of 3D data samples, while benefiting from the simplicity of 2D reconstructions. Our proposed multidirectional convolutional neural network has two parallel branches trained to perform 2D reconstructions along the vertical and horizontal directions and a small 3D part that combines their results. We use our method to reconstruct data gaps resulting from several missing shots in a benchmark synthetic and a real land dataset. Compared to a conventional 3D U-net, our network learns to reconstruct the events more accurately. Compared to 2D U-nets, our method avoids the discontinuities that arise from the 2D reconstruction of each trace of the missing shot gathers.
BIRD. BCAM's Institutional Repository Data
Publication Artículo científico (JournalArticle)

Solving boundary value problems via the Nyström method using spline Gauss rules

  • Hashemian, A.
  • Sliusarenko, H.
  • Remogna, S.
  • Barrera, D.
  • Barton, M.
We propose to use spline Gauss quadrature rules for solving boundary value problems (BVPs) using the Nyström method. When solving BVPs, one converts the corresponding partial differential equation inside a domain into the Fredholm integral equation of the second kind on the boundary in the sense of boundary integral equation (BIE). The Fredholm integral equation is then solved using the Nyström method, which involves the use of a particular quadrature rule, thus, converting the BIE problem to a linear system. We demonstrate this concept on the 2D Laplace problem over domains with smooth boundary as well as domains containing corners. We validate our approach on benchmark examples and the results indicate that, for a fixed number of quadrature points (i.e., the same computational effort), the spline Gauss quadratures return an approximation that is by one to two orders of magnitude more accurate compared to the solution obtained by traditional polynomial Gauss counterparts.
BIRD. BCAM's Institutional Repository Data
Publication Artículo científico (JournalArticle)

Predictive Maintenance of Floating Offshore Wind Turbine Mooring Lines using Deep Neural Networks

  • Gorostidi, N.
  • Nava, V.
  • Aristondo, A.
  • Pardo, D.
The recent massive deployment of onshore wind farms has caused controversy to arise mainly around the issues of land occupation, noise and visual pollution and impact on wildlife. Fixed offshore turbines, albeit beneficial in those aspects, become economically unfeasible when installed far away from coastlines. The possibility of installing floating offshore wind turbines is currently hindered by their excessive operation and maintenance costs. We have developed a comprehensive model to help companies plan their operations in advance by detecting failure in mooring lines in almost real time using supervised deep learning techniques. Given the lack of real data, we have coupled numerical methods and OpenFAST simulations to build a dataset containing the displacements and rotations of a turbine's floating platform across all directions. These time series and their corresponding frequency spectra are used to obtain a set of key statistical parameters, including means and standard deviations, peak frequencies, and several relevant momenta. We have designed and trained a Deep Neural Network to understand and distinguish amongst a series of common failure modes for mooring lines considering a range of metocean and structural conditions. We have obtained promising results when monitoring severe changes in the line's mass and damping using short time spans, achieving a 95.7\% validation accuracy when detecting severe biofouling failure.
BIRD. BCAM's Institutional Repository Data
Publication Artículo científico (JournalArticle)

Machine learning discovery of optimal quadrature rules for isogeometric analysis

  • Teijeiro, T.
  • Taylor, J.M.
  • Hashemian, A.
  • Pardo, D.
We propose the use of machine learning techniques to find optimal quadrature rules for the construction of stiffness and mass matrices in isogeometric analysis (IGA). We initially consider 1D spline spaces of arbitrary degree spanned over uniform and non-uniform knot sequences, and then the generated optimal rules are used for integration over higher-dimensional spaces using tensor products. The quadrature rule search is posed as an optimization problem and solved by a machine learning strategy based on adaptive gradient-descent. However, since the optimization space is highly non-convex, the success of the search strongly depends on the number of quadrature points and the parameter initialization. Thus, we use a dynamic programming strategy that initializes the parameters from the optimal solution over the spline space with a lower number of knots. With this method, we found optimal quadrature rules for spline spaces when using IGA discretizations with up to 50 uniform elements and polynomial degrees up to 8, showing the generality of the approach in this scenario. For non-uniform partitions, the method also finds an optimal rule in a reasonable number of test cases. We also assess the generated optimal rules in two practical case studies, namely, the eigenvalue problem of the Laplace operator and the eigenfrequency analysis of freeform curved beams, where the latter problem shows the applicability of the method to curved geometries. In particular, the proposed method results in savings with respect to traditional Gaussian integration of up to 44% in 1D, 68% in 2D, and 82% in 3D spaces., Euskampus Foundation through the ORLEG-IA project in the Misiones Euskampus 2.0 program. RYC2021-032853-I/MCIN/AEI/10.13039/501100011033 funded by the Spanish Ministry of Science and Innovation and by the European Union NextGenerationEU/PRTR .
BIRD. BCAM's Institutional Repository Data
Publication Artículo científico (JournalArticle)

Diagnosis of the health status of mooring systems for floating offshore wind turbines using autoencoders

  • Gorostidi, N.
  • Pardo, D.
  • Nava, V.
Floating offshore wind turbines (FOWTs) show promise in terms of energy production, availability, and sustainability, but remain unprofitable due to high maintenance costs. This work proposes a deep learning algorithm to detect mooring line degradation and failure by monitoring the dynamic response of the publicly available DeepCWind OC4 semi-submersible platform. This study implements an autoencoder capable of predicting multiple forms of damage occurring at once, with various levels of severity. Given the scarcity of real data, simulations performed in OpenFAST, recreating both healthy and damaged mooring systems, are used to train and validate the algorithm. The novelty of the proposed approach consists of using a set of key statistical metrics describing the platform’s displacements and rotations as input layer for the autoencoder. The statistics of the responses are calculated at 33-minute-long sea states under a broad spectrum of metocean and wind conditions. An autoencoder is trained using these parameters to discover that the proposed algorithm is capable of detecting mild anomalies caused by biofouling and anchor displacements, with correlation coefficients up to 98.51\% and 99.16\%, respectively. These results are encouraging for the continuous health monitoring of FOWT mooring systems using easily measurable quantities to plan preventive maintenance actions adequately.
BIRD. BCAM's Institutional Repository Data
Publication Capítulo o parte de un libro (BookPart)

Memory-Based Monte Carlo Integration for Solving Partial Differential Equations Using Neural Networks

  • Uriarte, C.
  • Taylor, J.M.
  • Pardo, D.
  • Rodríguez, O.A.
  • Vega
Monte Carlo integration is a widely used quadrature rule to solve Partial Differential Equations with neural networks due to its ability to guarantee overfitting-free solutions and high-dimensional scalability. However, this stochastic method produces noisy losses and gradients during training, which hinders a proper convergence diagnosis. Typically, this is overcome using an immense (disproportionate) amount of integration points, which deteriorates the training performance. This work proposes a memory-based Monte Carlo integration method that produces accurate integral approximations without requiring the high computational costs of processing large samples during training.
BIRD. BCAM's Institutional Repository Data
Publication Artículo científico (JournalArticle)

Multimodal variational autoencoder for inverse problems in geophysics: application to a 1-D magnetotelluric problem

  • Rodriguez, O.
  • Taylor, J.M.
  • Pardo, D.
Estimating subsurface properties from geophysical measurements is a common inverse problem. Several Bayesian methods currently aim to find the solution to a geophysical inverse problem and quantify its uncertainty. However, most geophysical applications exhibit more than one plausible solution. Here, we propose a multimodal variational autoencoder model that employs a mixture of truncated Gaussian densities to provide multiple solutions, along with their probability of occurrence and a quantification of their uncertainty. This autoencoder is assembled with an encoder and a decoder, where the first one provides a mixture of truncated Gaussian densities from a neural network, and the second is the numerical solution of the forward problem given by the geophysical approach. The proposed method is illustrated with a 1-D magnetotelluric inverse problem and recovers multiple plausible solutions with different uncertainty quantification maps and probabilities that are in agreement with known physical observations., PDC2021-121093-I00 IA4TES
BIRD. BCAM's Institutional Repository Data
Publication Artículo científico (JournalArticle)

Fast parallel IGA-ADS solver for time-dependent Maxwell's equations

  • Los, M.
  • Wozniak, M.
  • Pingali, K.
  • Garcia-Castillo, L.E.
  • Alvarez-Aramberri, J.
  • Pardo, D.
  • Paszyński, M.
We propose a simulator for time-dependent Maxwell's equations with linear computational cost. We employ B-spline basis functions as considered in the isogeometric analysis (IGA). We focus on non-stationary Maxwell's equations defined on a regular patch of elements. We employ the idea of alternating-directions splitting (ADS) and employ a second-order accurate time-integration scheme for the time-dependent Maxwell's equations in a weak form. After discretization, the resulting stiffness matrix exhibits a Kronecker product structure. Thus, it enables linear computational cost LU factorization. Additionally, we derive a formulation for absorbing boundary conditions (ABCs) suitable for direction splitting. We perform numerical simulations of the scattering problem (traveling pulse wave) to verify the ABC. We simulate the radiation of electromagnetic (EM) waves from the dipole antenna. We verify the order of the time integration scheme using a manufactured solution problem. We then simulate magnetotelluric measurements. Our simulator is implemented in a shared memory parallel machine, with the GALOIS library supporting the parallelization. We illustrate the parallel efficiency with strong and weak scalability tests corresponding to non-stationary Maxwell simulations., EXPERTIA (KK-2021/00048) SIGZE (KK-2021/00095) PDC2021-121093-I00
BIRD. BCAM's Institutional Repository Data
Publication Artículo científico (JournalArticle)

r-Adaptive deep learning method for solving partial di erential equations

  • Pardo, D.
  • Omella, A. J.
We introduce a Deep Neural Network (DNN) method for solving Partial Di erential Equations (PDEs) that simultaneously: (a) constructs an optimal r􀀀adapted mesh, i.e., given an initial mesh, it provides optimal node locations, and (b) solves the PDE over the constructed r-adaptive mesh. The node locations are optimized over a set of 1D boundary nodes, and the corresponding 2D quadrilateral meshes are built using tensor product. The method supports the definition of fixed interfaces to create conforming meshes and allows nodes to jump across them, permitting topological variations. To numerically illustrate the performance of our novel r􀀀adaptive deep learning method, we apply it in combination with other numerical methods including collocation, Least Squares, and the Deep Ritz method. We solve one- and two-dimensional problems whose solutions are smooth, singular, and/or exhibit strong gradients. Results consistently show the outperformance of employing r􀀀adaptivity, while in some cases the improvement is limited by the tensor-product structure of the mesh., IA4TES (MIA.2021.M04.008 / NextGenerationEU PRTR); PDC2021-121093-I00 (MCIN/ AEI / 10.13039/501100011033/Next Generation EU)
BIRD. BCAM's Institutional Repository Data
Publication Artículo científico (JournalArticle)

Condition monitoring of mooring systems for Floating Offshore Wind Turbines using Convolutional Neural Network framework coupled with Autoregressive coefficients

  • Sharma, S.
  • Nava, V.
This research presents a novel approach proposed for the monitoring of mooring systems in Floating Offshore Wind Turbines (FOWTs), employing a combination of Convolutional Neural Networks (CNNs) and Auto-Regressive (AR) models. CNN finds broad application in monitoring intricate structures, as they adeptly handle noisy response data without necessitating profound domain expertise. The precision of CNNs relies on the extraction of meaningful features from input data, necessitating meticulous data curation and labeling for optimal computational efficiency and accurate estimation. Emphasis is placed on the preference for feature-rich small datasets over voluminous yet sparse datasets, aiming to enable CNNs to discern crucial patterns more effectively and mitigate issues such as overfitting and extensive preprocessing. The novelty of the proposed approach lies in the integration of AR models, which serve to compress data and enhance damage-sensitive characteristics in the input for CNNs. This integration involves deploying regression models fitted to historical responses, parameterized with AR coefficients sensitive to damage, and further classifying severity using CNNs. The sequential nature of this approach addresses challenges such as vanishing/exploding gradients, particularly for extended historical data, while also attenuating the impact of noise and irrelevant information through data compression. The study explores the effectiveness of the coupled AR-CNN method in monitoring FOWT mooring lines, with a specific focus on two levels of damage identification: detection with classification and damage severity across diverse damage and operational scenarios. The modified methodology exhibits superior outcomes by conducting a performance analysis against traditional CNNs and other machine-learning methods, highlighting the potential of the AR-CNN strategy to improve the precision of FOWT mooring line condition monitoring. These findings underscore the AR-CNN strategy's potential to enhance the accuracy of FOWT mooring line condition monitoring.
BIRD. BCAM's Institutional Repository Data
Publication Artículo científico (JournalArticle)

Control co-design for wave energy farms: Optimisation of array layout and mooring configuration in a realistic wave climate

  • Peña-Sanchez, Y.
  • García-Violini, D.
  • Penalba, M.
  • Zarketa, A.
  • Ferri, F.
  • Nava, V.
  • Ringwood, J.V.
This paper presents a novel Control Co-Design (CCD) methodology aimed at economically optimising the layout of wave energy converter (WEC) arrays. CCD ensures the synergy of optimised WEC and array parameters with the final control strategy, resulting in a comprehensive and efficient design of the array. By integrating a spectral-based control strategy into the array layout design, this study pursues the twin objectives of maximising energy absorption while reducing costs. To prove the performance of the proposed CCD methodology, an application case is proposed where the inter-device distance, alignment, and mooring configuration of a five-device array, considering realistic wave scenarios, are optimised. Energy capture and system cost evaluations are conducted, with results emphasising the significance of incorporating advanced control strategies in the design phase to improve energy absorption and reduce costs. With the application case, the study demonstrates that the optimal layout of a WEC array considering economic factors may differ from the optimal from purely technical factors, such as energy absorption, in the analysed case.
BIRD. BCAM's Institutional Repository Data
Publication Artículo científico (JournalArticle)

Deep Neural Network for damage detection in Infante Dom Henrique bridge using multi-sensor data

  • Fernandez-Navamuel, A.
  • Pardo, D.
  • Magalhães, F.
  • Zamora-Sánchez, D.
  • Omella, A. J.
  • Garcia-Sanchez, D.
This paper proposes a data-driven approach to detect damage using monitoring data from the Infante Dom Henrique bridge in Porto. The main contribution of this work lies in exploiting the combination of raw measurements from local (inclinations and stresses) and global (eigenfrequencies) variables in a full-scale SHM application. We exhaustively analyze and compare the advantages and drawbacks of employing each variable type and explore the potential of combining them. An autoencoder-based Deep Neural Network is employed to properly reconstruct measurements under healthy conditions of the structure, which are influenced by environmental and operational variability. The damage-sensitive feature for outlier detection is the reconstruction error that measures the discrepancy between current and estimated measurements. Three autoencoder architectures are designed according to the input: local variables, global variables, and their combination. To test the performance of the methodology in detecting the presence of damage, we employ a Finite Element model to calculate the relative change in the structural response induced by damage at four locations. These relative variations between the healthy and damaged responses are employed to affect the experimental testing data, thus producing realistic time-domain damaged measurements. We analyze the Receiver Operating Curves and investigate the latent feature representation of the data provided by the autoencoder in the presence of damage. Results reveal the existence of synergies between the different variable types, producing almost perfect classifiers throughout the performed tests when combining the two available data sources. When damage occurs far from the instrumented sections, the area under the curve in the combined approach increases $50\%$ compared to using local variables only. The classificatoin metrics also demonstrate the enhancement of combining both sources of data in the damage detection task, reaching close to $97\%$ precission values for the four considered test damage scenarios. Finally, we also investigate the capability of local variables to localize the damage, demonstrating the potential of including these variables in the damage detection task., HAZITEK programme (ERROTAID project) and TCRINI project (KK-2023-0029) European Horizon (HE) with LIASON project (GA 101103698), and FUTURAL project (101083958)
BIRD. BCAM's Institutional Repository Data
Publication Artículo científico (JournalArticle)

Deep learning enhanced principal component analysis for structural health monitoring

  • Fernandez-Navamuel, A.
  • Magalhães, F.
  • Zamora-Sánchez, D.
  • Omella, A. J.
  • Garcia-Sanchez, D.
  • Pardo, D.
This paper proposes a Deep Learning enhanced Principal Component Analysis (PCA) approach for outlier detection to assess the structural condition of bridges. We employ a partially explainable autoencoder architecture to replicate and enhance the data compression and reconstruction ability of PCA. The particularity of the method lies in the addition of residual connections to account for nonlinearities. We apply the proposed method to monitoring data obtained from two bridges under real operation conditions and compare the results before and after adding the residual connections. Results show that the addition of residual connections enhances the outlier detection ability of the network, allowing to detect lighter damages., This work has received funding from: the European Union's Horizon 2020 research and innovation program under the grant agreement No 769373 (FORESEE project) and the Marie Sklodowska-Curie grant agreement No 777778 (MATHROCKS); the Base Funding - UIDB/04708/2020 of the CONSTRUCT - Instituto de I\&D em Estruturas e Construções - funded by national funds through the FCT/MCTES (PIDDAC); the European Regional Development Fund (ERDF) through the Interreg V-A Spain-France-Andorra program POCTEFA 2014-2020 Project PIXIL (EFA362/19); the Spanish Ministry of Science and Innovation with references PID2019-108111RB-I00 MCIN/AEI/10.13039/501100011033 (FEDER/AEI) and the “BCAM Severo Ochoa” accreditation of excellence (SEV-2017-0718); and the Basque Government through the BERC 2018-2021 program, the two Elkartek projects 3KIA (KK-2020/00049) and MATHEO (KK-2019-00085), the grant "Artificial Intelligence in BCAM number EXP. 2019/00432", and the Consolidated Research Group MATHMODE (IT1294-19) given by the Department of Education.
BIRD. BCAM's Institutional Repository Data
Publication Artículo científico (JournalArticle)

Performance measures of nonstationary inventory models for perishable products under the EWA policy

  • Pardo, D.
  • Gorria, C.
  • Lezaun, M.
Accurately estimating key performance indicators in inventory models for perishable items is essential in order to assess and improve the management strategy of these systems. We analyse the production of platelet concentrates at blood banks under the EWA replenishment policy. We give analytical approximations of the most important performance measures, such as the size of orders, the size of stocks, the percentage of outdating, the age distribution of stocks and the freshness of units issued, among others. The production of platelet concentrates is a prototypical example of inventory models for short life items with random demand and a weekly pattern, where a high service level is required. The methodology and the approximations presented here can be easily adapted to other inventory systems with similar characteristics. Most of the formulae in this article are new for nonstationary models under the EWA policy; indeed, formulae for the age distribution of units in stock and of units issued have not appeared in the literature even for the simpler base-stock replenishment policy. We apply our results to a real blood bank and find very close agreement between the formulae and the results of Monte Carlo simulations. The accuracy of our approximations is also tested in several scenarios, depending on the lifetime of units, safety stock levels and the probabilistic distribution of demand., PID2020-116873GB-I00
BIRD. BCAM's Institutional Repository Data
Publication Artículo científico (JournalArticle)

Development of methods based on neural networks in the estimation of mineral resources

  • Goti, A.
  • Alberdi, E.
  • Hernandez, H.
  • Goti, A.
Debido a limitaciones económicas y físicas, nuestra comprensión de los recursos minerales en un área de interés específica es limitada y fragmentada. Tradicionalmente, este problema se ha resuelto utilizando el método geoestadístico de Kriging, donde la ley del mineral se estima en ubicaciones sin mediciones utilizando valores conocidos de la ley en puntos circundantes. La ventaja de este método radica en el cálculo de pesos a través de un modelo de variabilidad espacial conocido como variograma. Sin embargo, su desventaja es que se basa en el supuesto de estacionariedad, aditividad, linealidad y de cierta forma el modelado variográfico es subjetivo. Este estudio propone abordar el problema de estimación de recursos minerales como un problema de regresión utilizando redes neuronales, que no están sujetas a las restricciones de estacionariedad, aditividad, linealidad y modelado espacial de los métodos geoestadísticos. Se han comparado el Kriging y una red neuronal de función de base radial y un perceptrón multicapa utilizando distintas métricas de validación. Los resultados muestran que un modelo de red neuronal adecuadamente entrenado, con un etiquetado apropiado de la ley mineral y sus características de entrada, logra resultados similares al enfoque geoestadístico con una reducción de tiempo significativo, evitando a su vez todos los supuestos antes señalados. Sin embargo, las redes neuronales no consideran la correlación espacial de la ley del mineral ni la reproducen en los lugares donde se midió, características que han marcado desconfianza en su implementación industrial y que se discuten en este artículo proponiendo finalmente un ajuste entre ambos enfoques a un mínimo sacrificio del coste temporal y de mano de obra.
BIRD. BCAM's Institutional Repository Data
Publication Tesis doctoral (DoctoralThesis)

Damage identification in bridges combining deep learning and computational mechanic

  • Fernandez-Navamuel, A.
Civil infrastructures, such as bridges, are critical assets for society and the economy. Many of them have already reached their expected life and withstand loadings that exceed the design specifications. Besides, bridges suffer from various degradation mechanisms, including aging, corrosion, earthquakes, and, nowadays, the undeniable effect of climate change. This context has motivated an increasing interest in early detecting damage to prevent costly actions and dangerous failures. Structural Health Monitoring (SHM) consists of implementing effective strategies to continuously assess the health condition of structures using monitoring data collected by sensors. This dissertation focuses on the SHM problem of damage detection and identification. It is an ill-posed inverse problem that aims at inferring the health state of a structure from measurements of its response. The measurements include large amounts of noisy data affected by environmental and operational conditions, acquired with sensors of different nature. Solving such a multidisciplinary problem encompasses the use of applied mathematics, computational mechanics, and data science. In this dissertation, we exploit the potential of Deep Neural Networks in approximating complex inverse problems and employ computational parametrizations and the Finite Element Method to enrich the training phase by including damage scenarios. We explore two different approaches to the problem. In the first approach, we develop an outlier detection strategy to detect departures from the baseline condition. We only employ long-term monitoring data measured at the bridge during normal (healthy) operation. Starting from Principal Component Analysis (PCA) as a statistical data reconstruction technique, we design a specific Deep Autoencoder network that enhances PCA by adding residual connections to include nonlinear transformations. This architecture gains partial explainability by evaluating the contribution of nonlinearties over affine transformations in the reconstruction process. We also investigate the method performance when using local or global variables and evaluate the potential of combining both data sources in the damage detection task. In the second approach, we reach a higher level of damage identification by estimating its severity and location. The goal is to provide a suitable methodology for real full-scale applications that requires reasonable computational resources. We employ a calibrated computational parametrization to solve multiple Finite Element simulations under different damage scenarios. These synthetic scenarios enrich the training dataset of a Deep Neural Network that maps the response of the bridge with its health condition in terms of damage location and severity. Finally, we incorporate the effect of environmental and operational variability in the parametrization by applying a clustering algorithm to find representative samples among the entire dataset. We assume these samples cover most of the variability present in the data and consider them as starting points to generate synthetic training data. We apply the proposed methods to three main case study bridges with available monitoring data: the Beltran bridge in Mexico, and the Infante Dom Henrique bridge in Porto, and the Z24 bridge in Switzerland. Both structures resulted critical to validate and test the ability of the proposed methods and to demonstrate their applicability in the full-scale., This disseration has been possible thanks to the support received from: the European Union’s Horizon 2020 research and innovation program under the grant agreement No 769373 (FORESEE project) and the Marie Sklodowska-Curie grant agreement No 777778 (MATHROCKS); the Base Funding - UIDB/04708/2020 of the CONSTRUCT - Instituto de I&D em Estruturas e Constru¸c˜oes - funded by national funds through the FCT/MCTES (PIDDAC); the European Regional Development Fund (ERDF) through the Interreg V-A Spain-France-Andorra program POCTEFA 2014-2020 Project PIXIL (EFA362/19); the Spanish Ministry of Science and Innovation with references PID2019-108111RB-I00 (FEDER/AEI) and the “BCAM Severo Ochoa” accreditation of excellence (SEV-2017-0718); and the Basque Government through the BERC 2018-2021 program, the four Elkartek projects 3KIA (KK-2020/00049), EXPERTIA (KK-2021/00048), MATHEO (KK-2019-00085), and SIGZE (KK-2021/00095); the grant “Artificial Intelligence in BCAM number EXP. 2019/00432”, and the Consolidated Research Group MATHMODE (IT1294-19) given by the Department of Education.
BIRD. BCAM's Institutional Repository Data
Publication Artículo científico (JournalArticle)

Reducing spatial discretization error on coarse CFD simulations using an openFOAM-embedded deep learning framework

  • Gonzalez-Sieiro, J.
  • Pardo, D.
  • Nava, V.
  • Calo, V.M.
  • Towara, M.
We propose a method for reducing the spatial discretization error of coarse computational fluid dynamics (CFD) problems by enhancing the quality of low-resolution simulations using deep learning. We feed the model with fine-grid data after projecting it to the coarse-grid discretization. We substitute the default differencing scheme for the convection term by a feed-forward neural network that interpolates velocities from cell centers to face values to produce velocities that approximate the down-sampled fine-grid data well. The deep learning framework incorporates the open-source CFD code OpenFOAM, resulting in an end-to-end differentiable model. We automatically differentiate the CFD physics using a discrete adjoint code version. We present a fast communication method between TensorFlow (Python) and OpenFOAM (c++) that accelerates the training process. We applied the model to the flow past a square cylinder problem, reducing the error from 120% to 25% in the velocity for simulations inside the training distribution compared to the traditional solver using an x8 coarser mesh. For simulations outside the training distribution, the error reduction in the velocities was about 50%. The training is affordable in terms of time and data samples since the architecture exploits the local features of the physics., PID2023-146678OB-I00 PRE2020-093093
BIRD. BCAM's Institutional Repository Data
Publication Artículo científico (JournalArticle)

Adaptive Deep Fourier Residual method via overlapping domain decomposition

  • Taylor, J.M.
  • Bastidas, M.
  • Calo, V. M.
  • Pardo, D.
The Deep Fourier Residual (DFR) method is a specific type of variational physics-informed neural network (VPINN). It provides a robust neural network-based solution to partial differential equations (PDEs). The DFR strategy is based on minimizing the dual norm of the weak residual of a PDE, which is equivalent to minimizing the energy norm of the error. To compute the dual norm of the weak residual, the DFR method employs an orthonormal spectral basis of the test space, known for rectangles or cuboids for multiple function spaces. In this work, we extend the DFR method with ideas of traditional domain decomposition (DD). This enables two improvements: (a) to solve problems in more general polygonal domains and (b) to develop an adaptive refinement technique in the test space using a Döfler marking algorithm. In the former case, we retain the desirable equivalence between the employed loss function and the H1 error under non-restrictive assumptions, numerically demonstrating adherence to explicit bounds in the case of the L-shaped domain problem. In the latter, we show how refinement strategies lead to potentially significant improvements against a reference, classical DFR implementation with a test function space of significantly lower dimensionality, allowing us to better approximate singular solutions at a more reasonable computational cost., PDC2021-121093-I00 (MCIN/AEI/10.13039/501100011033/Next Generation EU) Consolidated Research Group MATHMODE (IT1456-22)
BIRD. BCAM's Institutional Repository Data
Publication Artículo científico (JournalArticle)

Ensemble Deep Learning for Enhanced Seismic Data Reconstruction

  • Abedi, M.M.
  • Pardo, D.
  • Alkhalifah, T.
Seismic data often contain gaps due to various obstacles in the investigated area and recording instrument failures. Deep-learning techniques offer promising solutions for reconstructing missing data parts by utilizing existing data. Nonetheless, self-supervised methods frequently struggle with capturing under-represented features such as weaker events, crossing dips, and higher frequencies. To address these challenges, we propose a novel ensemble deep model (EDM) along with a tailored self-supervised training approach for reconstructing seismic data with consecutive missing traces. Our model comprises two branches of U-nets, each fed from distinct data transformation modules aimed at amplifying under-represented features and promoting diversity among learners. Our loss function minimizes relative errors at the outputs of individual branches and the entire model, ensuring accurate reconstruction of various features while maintaining overall data integrity. Additionally, we employ masking while training to enhance sample diversity and memory efficiency. Applications on two benchmark synthetic datasets and two real datasets demonstrate improved accuracy compared to a conventional U-net, successfully reconstructing weak events, diffractions, higher frequencies, and reflections covered by groundroll. Despite these advancements, our method does incur three times the training cost compared to a simple U-net.
BIRD. BCAM's Institutional Repository Data
Publication Artículo científico (JournalArticle)

Targeting the high frequency tail of wave spectra for energy harvesting in marine sensor networks

  • Davidson, J.P.
  • Nava, V.
While the conventional philosophy of wave energy conversion is to target the large amounts of power in the peak of the input wave spectrum, this study proposes that for the application of powering marine sensor networks (MSNs), it is advantageous to target the high frequency tail of the wave spectrum. This strategy is predicated on two primary advantages: the spatial and temporal persistence of the wave energy resource in the high-frequency region and its compatibility with the resonance characteristics of smaller MSN devices. The research seeks to identify the frequency range in sea spectra which is temporally and spatially omnipresent in the marine environment, and quantify the energy content available within this range. Theoretical frameworks are combined with a detailed case study to analyse the temporal and spatial variations of the wave resource and the persistent energy content in the high-frequency tail. Results reveal a notable consistency in energy availability above 2.5 rad/s, highlighting the high-frequency tail's potential as a reliable energy resource for MSNs. The available power in the identi ed high-frequency range is quanti ed, providing valuable insights for the design and implementation of e cient wave energy harvesters speci cally tailored to the needs of diverse marine environments., PRTR-C17.I1
BIRD. BCAM's Institutional Repository Data
Publication Artículo científico (JournalArticle)

Deep Fourier Residual method for solving time-harmonic Maxwell’s equations

  • Taylor, J.M.
  • Pardo, D.
  • Muga, I.
  • Bastidas, M.
Solving PDEs with machine learning techniques has become a popular alternative to conventional methods. In this context, Neural networks (NNs) are among the most commonly used machine learning tools, and in those models, the choice of an appropriate loss function is critical. In general, the main goal is to guarantee that minimizing the loss during training translates to minimizing the error in the solution at the same rate. In this work, we focus on the time-harmonic Maxwell’s equations, whose weak formulation takes H0(curl,Omega) as the space of test functions. We propose a NN in which the loss function is a computable approximation of the dual norm of the weak-form PDE residual. To that end, we employ the Helmholtz decomposition of the space H0(curl,Omega) and construct an orthonormal basis for this space in two and three spatial dimensions. Here, we use the Discrete Sine/Cosine Transform to accurately and efficiently compute the discrete version of our proposed loss function. Moreover, in the numerical examples we show a high correlation between the proposed loss function and the H(curl)-norm of the error, even in problems with low-regularity solutions., PDC2021-121093-I00
BIRD. BCAM's Institutional Repository Data
Publication Artículo científico (JournalArticle)

Development and Validation of a Health-aware Floating Offshore Wind Farm Simulation Platform: FOWLTY

  • Peña-Sanchez, Y.
  • Penalba, M.
  • Knudsen, T.
  • Nava, V.
  • Pardo, D.
Offshore wind energy is now an important game player towards the achievement of United Nations’ Sustainable Goals 7 (clean and affordable energy) and 13 (climate change) due to its huge potential but, given the harsh environment in which the turbines operate, they lead to more frequent and severe component faults compared to onshore farms. These faults result in increased maintenance and operational costs, posing a significant challenge to profitability. To mitigate these costs and maintain energy production during faults, fault detection and isolation (FDI) and fault-tolerant control (FTC) strategies have gained traction. Effective simulation tools are crucial for the development of these strategies. FOWLTY, a new simulator introduced in this paper, which derives its name from ’floating offshore wind farm simulator’ incorporating the term ’faulty’, allows users to recreate faults in wind turbine subsystems. It simplifies simulation to facilitate the development of control and estimation strategies, offering a balance between accuracy and computational efficiency. FOWLTY is compared to OpenFAST considering different scenarios, demonstrating its suitability for fault detection, diagnosis and impact mitigation through control. Additionally, a case study considering a farm composed of eight floating wind turbines is provided to show the effects of different faults., PDC2021-121093-I00
BIRD. BCAM's Institutional Repository Data
Publication Artículo científico (JournalArticle)

Fault detection and identification for control systems in floating offshore wind farms: A supervised Deep Learning methodology

  • Fernandez-Navamuel, A.
  • Peña-Sanchez, Y.
  • Nava, V.
This study employs a data-driven Fault Detection and Isolation (FDI) methodology in Floating Offshore Wind Turbine (FOWT) farms. The main objective of the work lies in classifying faults impacting the components of the control subsystems across multiple turbines. Unlike existing research, the emphasis here is placed specifically on identifying and classifying non-critical faults, which may result in suboptimal farm performance without necessitating a shutdown. From a methodological perspective, a Deep Neural Network has been designed to solve the classification problem by providing a probability vector, the most probable class indicator of the true state. One of the major contributions of this work lies in its applicability to FOWT farms instead of being confined to individual devices, facilitating a comprehensive performance assessment at the global farm level. The integration of this data-driven methodology with tolerant control strategies might enable early intervention, mitigating the impact of these faults and enhancing overall power generation efficiency. The target case study is a three-FOWT farm modeled in a Simulink environment, allowing for the simulation of operational behavior under diverse conditions and various faults affecting sensors and actuators. This work considers ten distinct fault classes, including the healthy condition, and three possible faults for each FOWT: pitch angle sensor, pitch angle actuator, and generator speed sensor. These frequent faults pose challenges to the optimal functioning of the control system managing the FOWTs. The outcomes highlight that the estimated probability of the healthy state serves as a robust indicator for detecting unknown faults. Results also demonstrate the adequate efficacy of the method in pinpointing the fault origin. However, we observe confusion between pitch sensor and actuator faults that require further investigation for comprehensive understanding., HAZITEK program (ERROTAID project (ZL-2022/00317)), ELKARTEK program (TCRINI project (KK-2023-0029) European Horizon (HE) with LIASON project (GA 101103698), FUTURAL project (101083958), and MsC Staff Exchange project RECHARGED (101086413). Yerai Peña-Sanchez is funded by the European Union’s Horizon 2020 research and innovation program under the Marie Sklodowska-Curie grant agreements N◦ 10103429
DIGITAL.CSIC. Repositorio Institucional del CSIC
Publication Artículo científico (JournalArticle)

Massive database generation for 2.5D borehole electromagnetic measurements using refined isogeometric analysis

  • Hashemian, Ali
  • García, Daniel
  • Rivera, Jon Ander
  • Pardo, David
Borehole resistivity measurements are routinely inverted in real-time during geosteering operations. The inversion process can be efficiently performed with the help of advanced artificial intelligence algorithms such as deep learning. These methods require a massive dataset that relates multiple Earth models with the corresponding borehole resistivity measurements. In here, we propose to use an advanced numerical method — refined isogeometric analysis (rIGA) — to perform rapid and accurate 2.5D simulations and generate databases when considering arbitrary 2D Earth models. Numerical results show that we can generate a meaningful synthetic database composed of 100,000 Earth models with the corresponding measurements in 56 h using a workstation equipped with two CPUs., This work has received funding from the European Union’s Horizon 2020 research and innovation programme under the Marie Sklodowska-Curie grant agreement No. 777778 (MATHROCKS), the European POCTEFA 2014–2020 Project PIXIL (EFA362/19) by the European Regional Development Fund (ERDF) through the Interreg V-A Spain-France-Andorra programme, the Project of the Spanish Ministry of Science and Innovation with reference PID2019-108111RB-I00 (FEDER/AEI), the BCAM “Severo Ochoa” accreditation of excellence (SEV-2017-0718), and the Basque Government through the BERC 2018–2021 programme, the two Elkartek projects 3KIA (KK-2020/00049) and MATHEO (KK-2019-00085), the grant “Artificial Intelligence in BCAM number EXP. 2019/00432”, and the Consolidated Research Group MATHMODE (IT1294-19) given by the Department of Education. The authors also acknowledge the computer resources and the technical support provided by Barcelona Supercomputing Center through the “MareNostrum4” cluster and under the activity ID “IM-2020-3-0009”., Peer reviewed
Zaguán. Repositorio Digital de la Universidad de Zaragoza
Publication Artículo científico (JournalArticle)

Sensitivity and uncertainty analysis by discontinuous Galerkin of lock-in thermography for crack characterization

  • Omella, Á.J.
  • Celorrio, R.
  • Pardo, D.
This work focuses on the characterization of narrow vertical cracks of finite size using optically excited lock-in thermography (OLT). To characterize these cracks, we need to solve an ill-posed inverse problem. As a previous step to the solution of this inverse problem, we propose a sensitivity analysis to quantify the influence that the parameters involved in the model have on the surface temperature. Some of these parameters are estimated at the laboratory and they incorporate uncertainty that may severely affect the reconstruction of thin cracks. For this reason, we design a calibration criterion based on the sensitivity analysis to determine which parameters we need to include as unknowns of the inverse problem. We perform this analysis using a numerical discontinuous Galerkin method. Additionally, we propose a theoretical noise model for the thermograms. Then, we use a weighted least square method (WLS) to determine the parameters from the experimental thermograms. We also obtain a theoretical uncertainty of the reconstructed parameters in OLT-WLS fitting according to the used surface temperature dataset. Finally, we perform a numerical experiment with a 2.4 m-thick vertical crack to show the sensitivities of the surface temperature with respect to the model parameters. We also determine the uncertainty of the parameters under different datasets with known noise characteristics.