Pasar al contenido principal

Resultados relacionados

Resultados totales (Incluyendo duplicados): 22
Encontrada(s) 1 página(s)

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

Neural network architecture optimization using automated machine learning for borehole resistivity measurements

  • Shahriari, M.
  • Pardo, D.
  • Kargaran, S.
  • Teijeiro, T.
Deep neural networks (DNNs) offer a real-time solution for the inversion of borehole resistivity measurements to approximate forward and inverse operators. Using extremely large DNNs to approximate the operators is possible, but it demands considerable training time. Moreover, evaluating the network after training also requires a significant amount of memory and processing power. In addition, we may overfit the model. In this work, we propose a scoring function that accounts for the accuracy and size of the DNNs compared to a reference DNNs that provides good approximations for the operators. Using this scoring function, we use DNN architecture search algorithms to obtain a quasi-optimal DNN smaller than the reference network; hence, it requires less computational effort during training and evaluation. The quasi-optimal DNN delivers comparable accuracy to the original large DNN., PDC2021-121093-I00 IA4TES RYC2021-032853-I
BIRD. BCAM's Institutional Repository Data
Publicación 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
Publicación 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
Publicación 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
Publicación 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
Publicación 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
Publicación 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
Publicación 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
Publicación 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
Publicación Artículo científico (JournalArticle)

A Least-Squares-Based Neural Network (LS-Net) for Solving Linear Parametric PDEs

  • Baharlouei, S.
  • Taylor, J.M.
  • Uriarte, C.
  • Pardo, D.
Developing efficient methods for solving parametric partial differential equations is crucial for addressing inverse problems. This work introduces a Least-Squares-based Neural Network (LS-Net) method for solving linear parametric PDEs. It utilizes a separated representation form for the parametric PDE solution via a deep neural network and a least-squares solver. In this approach, the output of the deep neural network consists of a vector-valued function, interpreted as basis functions for the parametric solution space, and the least-squares solver determines the optimal solution within the constructed solution space for each given parameter. The LS-Net method requires a quadratic loss function for the least-squares solver to find optimal solutions given the set of basis functions. In this study, we consider loss functions derived from the Deep Fourier Residual and Physics-Informed Neural Networks approaches. We also provide theoretical results similar to the Universal Approximation Theorem, stating that there exists a sufficiently large neural network that can theoretically approximate solutions of parametric PDEs with the desired accuracy. We illustrate the LS-net method by solving one- and two-dimensional problems. Numerical results clearly demonstrate the method's ability to approximate parametric solutions.
BIRD. BCAM's Institutional Repository Data
Publicación 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
Publicación Artículo científico (JournalArticle)

Learning quantities of interest from parametric PDEs: An efficient neural-weighted Minimal Residual approach

  • Brevis, R.I.
  • Muga, I.
  • Pardo, D.
  • Rodríguez, O.A.
  • Van der Zee, K.G.
The efficient approximation of parametric PDEs is of tremendous importance in science and engineering. In this paper, we show how one can train Galerkin discretizations to efficiently learn quantities of interest of solutions to a parametric PDE. The central component in our approach is an efficient neural-network-weighted Minimal-Residual formulation, which, after training, provides Galerkin-based approximations in standard discrete spaces that have accurate quantities of interest, regardless of the coarseness of the discrete space., PDC2021-121093-I00
BIRD. BCAM's Institutional Repository Data
Publicación 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
Publicación 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
Publicación Artículo científico (JournalArticle)

Gaussian Mixture autoencoder for uncertainty-aware damage identification in a Floating Offshore Wind Turbine

  • Fernandez-Navamuel, A.
  • Gorostidi, N.
  • Pardo, D.
  • Nava, V.
  • Chatzi, E.
This work proposes an uncertainty-aware approach to the inverse problem of damage identification in a Floating Offshore Wind Turbine (FOWT). We design an autoencoder architecture, where the latent space represents the features of the target damaged condition. The inverse operator (encoder) is a Deep Neural Network that maps the measurable response to the parameters (means, variances, and weights) of a multivariate Gaussian Mixture model. The Gaussian Mixture model provides a convenient distributional description that is flexible enough to accommodate complex solution spaces. The decoder receives samples from the Gaussian Mixture and maps the damaged condition (states) to the system’s measurable response. In such a problem, and depending on the quantities being observed (sensor positioning), it is possible that multiple damaged states may correspond to similar measurement records. In this context, the main contribution of this work lies in developing a method to quantify the uncertainty within the context of a possibly ill-posed damage identification problem. We employ the Gaussian Mixture to express the multimodal solution space and explain the uncertainty in the damaged condition estimates. We design and validate the methodology using synthetic data from a FOWT in the commonly adopted OpenFAST software and consider two damage types frequently occurring in mooring lines: biofouling and anchor displacement. The method allows for estimating the damaged state while capturing the uncertainty in the estimations and the multimodality of the solution under the availability of a limited number of response measurements., PID2023-146678OB-I00
BIRD. BCAM's Institutional Repository Data
Publicación 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
Publicación Artículo científico (JournalArticle)

Optimizing Variational Physics-Informed Neural Networks Using Least Squares

  • Uriarte, C.
  • Bastidas, M.
  • Pardo, D.
  • Taylor, J.M.
  • Rojas, S.
Variational Physics-Informed Neural Networks often suffer from poor convergence when using stochastic gradient-descent-based optimizers. By introducing a least squares solver for the weights of the last layer of the neural network, we improve the convergence of the loss during training in most practical scenarios. This work analyzes the computational cost of the resulting hybrid least-squares/gradient-descent optimizer and explains how to implement it efficiently. In particular, we show that a traditional implementation based on backward-mode automatic differentiation leads to a prohibitively expensive algorithm. To remedy this, we propose using either forward-mode automatic differentiation or an ultraweak-type scheme that avoids the differentiation of trial functions in the discrete weak formulation. The proposed alternatives are up to one hundred times faster than the traditional one, recovering a computational cost-per-iteration similar to that of a conventional gradient-descent-based optimizer alone. To support our analysis, we derive computational estimates and conduct numerical experiments in one- and two-dimensional problems.
BIRD. BCAM's Institutional Repository Data
Publicación 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