arXiv subjects

Esteban Ferrer

Publications and source records attributed to Esteban Ferrer.

At least 37 records · Page 2Linked to original sources

arXiv2024

Deep Reinforcement Learning for Multi-Objective Optimization: Enhancing Wind Turbine Energy Generation while Mitigating Noise Emissions

We develop a torque-pitch control framework using deep reinforcement learning for wind turbines to optimize the generation of wind turbine energy while minimizing operational noise. We employ a double deep Q-learning, coupled to a blade element momentum solver, to enable precise control over wind turbine parameters. In addition to the blade element momentum, we use the wind turbine acoustic model of Brooks Pope and Marcolini. Through training with simple winds, the agent learns optimal control policies that allow efficient control for complex turbulent winds. Our experiments demonstrate that the reinforcement learning is able to find optima at the Pareto front, when maximizing energy while minimizing noise. In addition, the adaptability of the reinforcement learning agent to changing turbulent wind conditions, underscores its efficacy for real-world applications. We validate the methodology using a SWT2.3-93 wind turbine with a rated power of 2.3 MW. We compare the reinforcement learning control to classic controls to show that they are comparable when not taking into account noise emissions. When including a maximum limit of 45 dB to the noise produced (100 meters downwind of the turbine), the extracted yearly energy decreases by 22%. The methodology is flexible and allows for easy tuning of the objectives and constraints through the reward definitions, resulting in a flexible multi-objective optimization framework for wind turbine control. Overall, our findings highlight the potential of RL-based control strategies to improve wind turbine efficiency while mitigating noise pollution, thus advancing sustainable energy generation technologies

eess.SY↗

arXiv2024

A reinforcement learning strategy to automate and accelerate h/p-multigrid solvers

We explore a reinforcement learning strategy to automate and accelerate h/p-multigrid methods in high-order solvers. Multigrid methods are very efficient but require fine-tuning of numerical parameters, such as the number of smoothing sweeps per level and the correction fraction (i.e., proportion of the corrected solution that is transferred from a coarser grid to a finer grid). The objective of this paper is to use a proximal policy optimization algorithm to automatically tune the multigrid parameters and, by doing so, improve stability and efficiency of the h/p-multigrid strategy. Our findings reveal that the proposed reinforcement learning h/p-multigrid approach significantly accelerates and improves the robustness of steady-state simulations for one dimensional advection-diffusion and nonlinear Burgers' equations, when discretized using high-order h/p methods, on uniform and nonuniform grids.

cs.LG↗

arXiv2024

Modelling Wind Turbines via Actuator Lines in High-Order h/p Solvers

This paper compares two actuator line methodologies for modelling wind turbines employing high-order h/p solvers and large-eddy simulations. The methods combine the accuracy of high-order solvers (in this work the maximum order is 6) with the computational efficiency of actuator lines to capture the aerodynamic effects of wind turbine blades. Comparisons with experiments validate the actuator line methodologies. We explore the effects of the polynomial order and the smoothing parameter associated with the Gaussian regularization function, and use them to blend the actuator line forcing in the high-order computational mesh, to show that both parameters influence the distribution of forces along the blades and the turbine wake. The greatest impact is obtained when the polynomial order is increased, allowing one to better capture the physics without requiring new meshes. When comparing the actuator line methodologies, we show the advantages of performing weighted sums, over element averages, to compute the blade velocities and forces in high-order solvers. For low-order simulations (low polynomial orders), both methods provide similar results, but as the polynomial order is increased, the weighted sum method shows smoother thrust/torque distributions and better wake resolution. Furthermore, cell averaging introduces nonphysical oscillations when increasing the polynomial order beyond 3 (4th order accuracy).

physics.flu-dyn↗

arXiv2024

Low-cost wind turbine aeroacoustic predictions using actuator lines

Aerodynamic noise is a limitation for further exploitation of wind energy resources. As this type of noise is caused by the interaction of turbulent flow with the airframe, a detailed resolution of the flow is necessary to obtain an accurate prediction of the far-field noise. Computational fluid dynamic (CFD) solvers simulate the flow field but only at a high computational cost, which is much increased when the acoustic field is resolved. Therefore, wind turbine noise predictions using numerical approaches remain a challenge. This paper presents a methodology that couples (relatively fast) wind turbine CFD simulations using actuator lines with a fast turn-around noise prediction method. The flow field is simulated using actuator lines and large eddy simulations. The noise prediction method is based on the Amiet-Schlinker's theory for rotatory noise sources, considering leading- and trailing-edge noise as unique noise sources. A 2D panel code (XFOIL) calculates the sectional forces and boundary layer quantities. The resulting methodology for the noise prediction method is of high fidelity since the wind turbine geometry is accounted for in both flow and acoustics predictions. Results are compared with field measurements of a full-scale wind turbine for two operational conditions, validating the results of this research.

physics.flu-dyn↗

arXiv2024

Accelerating high order discontinuous Galerkin solvers through a clustering-based viscous/turbulent-inviscid domain decomposition

We explore the unsupervised clustering technique introduced in [25] to identify viscous/turbulent from inviscid regions in incompressible flows. The separation of regions allows solving the Navier-Stokes equations including Large Eddy Simulation closure models only in the viscous/turbulent ones, while solving the Euler equations in the remaining of the computational domain. By solving different sets of equations, the computational cost is significantly reduced. This coupling strategy is implemented within a discontinuous Galerkin numerical framework, which allows discontinuous solutions (i.e., different set of equations) in neighbouring elements that interact through numerical fluxes. The proposed strategy maintains the same accuracy at lower cost, when compared to solving the full Navier-Stokes equations throughout the entire domain. Validation of this approach is conducted across diverse flow regimes, spanning from unsteady laminar flows to unsteady turbulent flows, including an airfoil section at Reynolds numbers Re = 103 and 104 and large angles of attack, and the flow past a wind turbine, modelled using actuator lines. The computational cost is reduced by 25% and 29% for the unsteady turbulent flow around an airfoil section and the flow past the wind turbine, respectively. In addition, to further accelerate the simulations, we combine the proposed decoupling with local P -adaptation, as proposed in [ 30]. When doing so, we reduce the computational cost by 41% and 45% for the flow around the airfoil section and the flow past the wind turbine, respectively

physics.flu-dyn↗

arXiv2024

Reinforcement learning to maximise wind turbine energy generation

We propose a reinforcement learning strategy to control wind turbine energy generation by actively changing the rotor speed, the rotor yaw angle and the blade pitch angle. A double deep Q-learning with a prioritized experience replay agent is coupled with a blade element momentum model and is trained to allow control for changing winds. The agent is trained to decide the best control (speed, yaw, pitch) for simple steady winds and is subsequently challenged with real dynamic turbulent winds, showing good performance. The double deep Q- learning is compared with a classic value iteration reinforcement learning control and both strategies outperform a classic PID control in all environments. Furthermore, the reinforcement learning approach is well suited to changing environments including turbulent/gusty winds, showing great adaptability. Finally, we compare all control strategies with real winds and compute the annual energy production. In this case, the double deep Q-learning algorithm also outperforms classic methodologies.

cs.LG↗

arXiv2024

A comparative study of explicit and implicit Large Eddy Simulations using a high-order discontinuous Galerkin solver: application to a Formula 1 front wing

This paper explores two Large Eddy Simulation (LES) approaches within the framework of the high-order discontinuous Galerkin solver, Horses3D. The investigation focuses on an Inverted Multi-element Wing in Ground Effect (i.e. 2.5D Imperial Front Wing section) representing a Formula 1 front wing, and compares the strengths and limitations of the two LES methods. The explicit LES formulation relies on the Vreman model, that adapts to laminar, transitional and turbulent regimes. The numerical formulation uses nodal basis functions and Gauss points. The implicit LES formulation, does not require explicit turbulence modeling but relies in the discretization scheme. We use the Kennedy-Gruber entropy stable formulation to enhance stability in under resolved simulations, since we recover the continuous properties such as entropy conservation at a discrete level. This formulation employs Gauss-Lobatto points, which downgrades the accuracy of integration but allows for larger time steps in explicit time integration. We compare our results to Nektar++ [1] showing that both LES techniques provide results that agree well with the reference values. The implicit LES shows to better capture transition and allows for larger time steps at a similar cost per iteration. We conclude that this implicit LES formulation is very attractive for complex simulations.

math.NA↗

arXiv2023

An unsupervised machine-learning-based shock sensor for high-order supersonic flow solvers

We present a novel unsupervised machine-learning sock sensor based on Gaussian Mixture Models (GMMs). The proposed GMM sensor demonstrates remarkable accuracy in detecting shocks and is robust across diverse test cases with significantly less parameter tuning than other options. We compare the GMM-based sensor with state-of-the-art alternatives. All methods are integrated into a high-order compressible discontinuous Galerkin solver, where two stabilization approaches are coupled to the sensor to provide examples of possible applications. The Sedov blast and double Mach reflection cases demonstrate that our proposed sensor can enhance hybrid sub-cell flux-differencing formulations by providing accurate information of the nodes that require low-order blending. Besides, supersonic test cases including high Reynolds numbers showcase the sensor performance when used to introduce entropy-stable artificial viscosity to capture shocks, demonstrating the same effectiveness as fine-tuned state-of-the-art sensors. The adaptive nature and ability to function without extensive training datasets make this GMM-based sensor suitable for complex geometries and varied flow configurations. Our study reveals the potential of unsupervised machine-learning methods, exemplified by this GMM sensor, to improve the robustness and efficiency of advanced CFD codes.

cs.LG↗

arXiv2023

A reinforcement learning strategy for p-adaptation in high order solvers

Reinforcement learning (RL) has emerged as a promising approach to automating decision processes. This paper explores the application of RL techniques to optimise the polynomial order in the computational mesh when using high-order solvers. Mesh adaptation plays a crucial role in improving the efficiency of numerical simulations by improving accuracy while reducing the cost. Here, actor-critic RL models based on Proximal Policy Optimization offer a data-driven approach for agents to learn optimal mesh modifications based on evolving conditions. The paper provides a strategy for p-adaptation in high-order solvers and includes insights into the main aspects of RL-based mesh adaptation, including the formulation of appropriate reward structures and the interaction between the RL agent and the simulation environment. We discuss the impact of RL-based mesh p-adaptation on computational efficiency and accuracy. We test the RL p-adaptation strategy on a 1D inviscid Burgers' equation to demonstrate the effectiveness of the strategy. The RL strategy reduces the computational cost and improves accuracy over uniform adaptation, while minimising human intervention.

cs.LG↗

arXiv2023

A combined volume penalization / selective frequency damping approach for immersed boundary methods: application to moving geometries

This work extends, to moving geometries, the immersed boundary method based on volume penalization and selective frequency damping approach [J. Kou, E. Ferrer, A combined volume penalization/selective frequency damping approach for immersed boundary methods applied to high-order schemes, Journal of Computational Physics (2023)]. To do so, the numerical solution inside the solid is decomposed into a predefined movement and an oscillatory part (spurious waves), where the latter is damped by an SFD approach combined with volume penalization. We challenge the method with two cases. First, a new manufactured solution problem is proposed to show that the method can recover high-order accuracy. Second, we validate the methodology by simulating the laminar flow past a moving cylinder, where improved accuracy of the combined method is reported.

physics.flu-dyn↗

arXiv2022

A modified equation analysis for immersed boundary methods based on volume penalization: applications to linear advection-diffusion and high-order discontinuous Galerkin schemes

The Immersed Boundary Method (IBM) is a popular numerical approach to impose boundary conditions without relying on body-fitted grids, thus reducing the costly effort of mesh generation. To obtain enhanced accuracy, IBM can be combined with high-order methods (e.g., discontinuous Galerkin). For this combination to be effective, an analysis of the numerical errors is essential. In this work, we apply, for the first time, a modified equation analysis to the combination of IBM (based on volume penalization) and high-order methods (based on nodal discontinuous Galerkin methods) to analyze a priori numerical errors and obtain practical guidelines on the selection of IBM parameters. The analysis is performed on a linear advection-diffusion equation with Dirichlet boundary conditions. Three ways to penalize the immerse boundary are considered, the first penalizes the solution inside the IBM region (classic approach), whilst the second and third penalize the first and second derivatives of the solution. We find optimal combinations of the penalization parameters, including the first and second penalizing derivatives, resulting in minimum errors. We validate the theoretical analysis with numerical experiments for one- and two-dimensional advection-diffusion equations.

math.NA↗

arXiv2022

Improving aircraft performance using machine learning: a review

This review covers the new developments in machine learning (ML) that are impacting the multi-disciplinary area of aerospace engineering, including fundamental fluid dynamics (experimental and numerical), aerodynamics, acoustics, combustion and structural health monitoring. We review the state of the art, gathering the advantages and challenges of ML methods across different aerospace disciplines and provide our view on future opportunities. The basic concepts and the most relevant strategies for ML are presented together with the most relevant applications in aerospace engineering, revealing that ML is improving aircraft performance and that these techniques will have a large impact in the near future.

cs.LG↗

arXiv2022

Truncation Error-Based Anisotropic $p$-Adaptation for Unsteady Flows for High-Order Discontinuous Galerkin Methods

In this work, we extend the $τ$-estimation method to unsteady problems and use it to adapt the polynomial degree for high-order discontinuous Galerkin simulations of unsteady flows. The adaptation is local and anisotropic and allows capturing relevant unsteady flow features while enhancing the accuracy of time evolving functionals (e.g., lift, drag). To achieve an efficient and unsteady truncation error-based $p$-adaptation scheme, we first revisit the definition of the truncation error, studying the effect of the treatment of the mass matrix arising from the temporal term. Secondly, we extend the $τ$-estimation strategy to unsteady problems. Finally, we present and compare two adaptation strategies for unsteady problems: the dynamic and static $p$-adaptation methods. In the first one (dynamic) the error is measured periodically during a simulation and the polynomial degree is adapted immediately after every estimation procedure. In the second one (static) the error is also measured periodically, but only one $p$-adaptation process is performed after several estimation stages, using a combination of the periodic error measures. The static $p$-adaptation strategy is suitable for time-periodic flows, while the dynamic one can be generalized to any flow evolution. We consider two test cases to evaluate the efficiency of the proposed $p$-adaptation strategies. The first one considers the compressible Euler equations to simulate the advection of a density pulse. The second one solves the compressible Navier-Stokes equations to simulate the flow around a cylinder at Re=100. The local and anisotropic adaptation enables significant reductions in the number of degrees of freedom with respect to uniform refinement, leading to speed-ups of up to $\times4.5$ for the Euler test case and $\times2.2$ for the Navier-Stokes test case.

math.NA↗

arXiv2022

Aeroacoustic airfoil shape optimization enhanced by autoencoders

We present a framework for airfoil shape optimization to reduce the trailing edge noise for the design of wind turbine blades. Far-field noise is evaluated using Amiet's theory coupled with the TNO-Blake model to calculate the wall pressure spectrum and fast turn-around XFOIL simulations to evaluate the boundary layer parameters. The computational framework is first validated using a NACA0012 airfoil at zero angle of attack. Particle swarm optimization is used to find the optimized airfoil configuration. The multi-objective optimization minimizes the A-weighted overall sound pressure level at various angles of attack, while ensuring enough lift and minimum drag. We compare classic parametrization methods to define the airfoil geometry (i.e., CST) to a machine learning method (i.e., a variational autoencoder). We observe that variational autoencoders can represent a wide variety of geometries, with only four encoded variables, leading to efficient optimizations, which result in improved optimal shapes. When compared to the baseline geometry, a NACA0012, the autoencoder-based optimized airfoil reduces by 3% (1.75 dBA) the overall sound pressure level (with decreased noise across the entire frequency range), while maintaining favorable aerodynamic properties in terms of lift and drag.

physics.flu-dyn↗

arXiv2022

Machine learning adaptation for laminar and turbulent flows: applications to high order discontinuous Galerkin solvers

We present a machine learning-based mesh refinement technique for steady and unsteady flows. The clustering technique proposed by Otmani et al. arXiv:2207.02929 [physics.flu-dyn] is used to mark the viscous and turbulent regions for the flow past a cylinder at Re=40 (steady laminar flow) and Re=3900 (unsteady turbulent flow). Within this clustered region, we increase the polynomial order to show that it is possible to obtain similar levels of accuracy to a uniformly refined mesh. The method is effective as the clustering successfully identifies the two flow regions, a viscous/turbulent dominated region (including the boundary layer and wake) and an inviscid/irrotational region (a potential flow region). The data used within this framework are generated using a high-order discontinuous Galerkin solver, allowing to locally refine the polynomial order (p-refinement) in each element of the clustered region. For the steady laminar test case we are able to reduce the computational cost up to 32% and for the unsteady turbulent case up to 33%.

physics.flu-dyn↗

arXiv2022

Jump penalty stabilisation techniques for under-resolved turbulence in discontinuous Galerkin schemes

Jump penalty stabilisation techniques have been recently proposed for continuous and discontinuous high order Galerkin schemes [1,2,3]. The stabilisation relies on the gradient or solution discontinuity at element interfaces to incorporate localised numerical diffusion in the numerical scheme. This diffusion acts as an implicit subgrid model and stablises under-resolved turbulent simulations. This paper investigates the effect of jump penalty stabilisation methods (penalising gradient or solution) for stabilisation and improvement of high-order discontinuous Galerkin schemes in turbulent regime. We analyse these schemes using an eigensolution analysis, a 1D non-linear Burgers equation (mimicking a turbulent cascade) and 3D turbulent Navier-Stokes simulations (Taylor-Green Vortex problem). We show that the two jump penalty stabilisation techniques can stabilise under-resolved simulations thanks to the improved dispersion-dissipation characteristics (when compared to non-penalised schemes) and provide accurate results for turbulent flows. The numerical results indicate that the proposed jump penalty stabilise under-resolved simulations and improve the simulations, when compared to the original unpenalised scheme and to classic explicit subgrid models (Smagorisnky and Vreman).

physics.flu-dyn↗

arXiv2022

Accelerating high order discontinuous Galerkin solvers using neural networks: 3D compressible Navier-Stokes equations

We propose to accelerate a high order discontinuous Galerkin solver using neural networks. We include a corrective forcing to a low polynomial order simulation to enhance its accuracy. The forcing is obtained by training a deep fully connected neural network, using a high polynomial order simulation but only for a short time frame. With this corrective forcing, we can run the low polynomial order simulation faster (with large time steps and low cost per time step) while improving its accuracy. We explored this idea for a 1D Burgers' equation in (Marique and Ferrer, CAF 2022), and we have extended this work to the 3D Navier-Stokes equations, with and without a Large Eddy Simulation closure model. We test the methodology with the turbulent Taylor Green Vortex case and for various Reynolds numbers (30, 200 and 1600). In addition, the Taylor Green Vortex evolves with time and covers laminar, transitional, and turbulent regimes, as time progresses. The proposed methodology proves to be applicable to a variety of flows and regimes. The results show that the corrective forcing is effective in all Reynolds numbers and time frames (excluding the initial flow development). We can train the corrective forcing with a polynomial order of 8, to increase the accuracy of simulations from a polynomial order 3 to 6, when correcting outside the training time frame. The low order correct solution is 4 to 5 times faster than a simulation with comparable accuracy (polynomial order 6). Additionally, we explore changes in the hyperparameters and use transfer learning to speed up the training. We observe that it is not useful to train a corrective forcing using a different flow condition. However, an already trained corrective forcing can be used to initialise a new training (at the correct flow conditions) to obtain an effective forcing with only a few training iterations.

physics.flu-dyn↗

arXiv2022

towards a robust detection of viscous and turbulent flow regions using unsupervised machine learning

We propose an invariant feature space for the detection of viscous dominated and turbulent regions (i.e., boundary layers and wakes). The developed methodology uses the principal invariants of the strain and rotational rate tensors as input to an unsupervised Machine Learning Gaussian mixture model. The selected feature space is independent of the coordinate frame used to generate the processed data, as it relies on the principal invariants of strain and rotational rate, which are Galilean invariants. This methodology allows us to identify two distinct flow regions: a viscous dominated, rotational region (boundary layer and wake region) and an inviscid, irrotational region (outer flow region). We test the methodology on a laminar and a turbulent (using Large Eddy Simulation) case for flows past a circular cylinder at $Re=40$ and $Re=3900$. The simulations have been conducted using a high-order nodal Discontinuous Galerkin Spectral Element Method (DGSEM). The results obtained are analysed to show that Gaussian mixture clustering provides an effective identification method of viscous dominated and rotational regions in the flow. We also include comparisons with traditional sensors to show that the proposed clustering does not depend on the selection of an arbitrary threshold, as required when using traditional sensors.

physics.flu-dyn↗