-
Augmenting a Physics-Informed Neural Network for the 2D Burgers Equation by Addition of Solution Data Points
Authors:
Marlon Sproesser Mathias,
Wesley Pereira de Almeida,
Marcel Rodrigues de Barros,
Jefferson Fialho Coelho,
Lucas Palmiro de Freitas,
Felipe Marino Moreno,
Caio Fabricio Deberaldini Netto,
Fabio Gagliardi Cozman,
Anna Helena Reali Costa,
Eduardo Aoun Tannuri,
Edson Satoshi Gomi,
Marcelo Dottori
Abstract:
We implement a Physics-Informed Neural Network (PINN) for solving the two-dimensional Burgers equations. This type of model can be trained with no previous knowledge of the solution; instead, it relies on evaluating the governing equations of the system in points of the physical domain. It is also possible to use points with a known solution during training. In this paper, we compare PINNs trained…
▽ More
We implement a Physics-Informed Neural Network (PINN) for solving the two-dimensional Burgers equations. This type of model can be trained with no previous knowledge of the solution; instead, it relies on evaluating the governing equations of the system in points of the physical domain. It is also possible to use points with a known solution during training. In this paper, we compare PINNs trained with different amounts of governing equation evaluation points and known solution points. Comparing models that were trained purely with known solution points to those that have also used the governing equations, we observe an improvement in the overall observance of the underlying physics in the latter. We also investigate how changing the number of each type of point affects the resulting models differently. Finally, we argue that the addition of the governing equations during training may provide a way to improve the overall performance of the model without relying on additional data, which is especially important for situations where the number of known solution points is limited.
△ Less
Submitted 18 January, 2023;
originally announced January 2023.
-
A Physics-Informed Neural Network to Model Port Channels
Authors:
Marlon S. Mathias,
Marcel R. de Barros,
Jefferson F. Coelho,
Lucas P. de Freitas,
Felipe M. Moreno,
Caio F. D. Netto,
Fabio G. Cozman,
Anna H. R. Costa,
Eduardo A. Tannuri,
Edson S. Gomi,
Marcelo Dottori
Abstract:
We describe a Physics-Informed Neural Network (PINN) that simulates the flow induced by the astronomical tide in a synthetic port channel, with dimensions based on the Santos - São Vicente - Bertioga Estuarine System. PINN models aim to combine the knowledge of physical systems and data-driven machine learning models. This is done by training a neural network to minimize the residuals of the gover…
▽ More
We describe a Physics-Informed Neural Network (PINN) that simulates the flow induced by the astronomical tide in a synthetic port channel, with dimensions based on the Santos - São Vicente - Bertioga Estuarine System. PINN models aim to combine the knowledge of physical systems and data-driven machine learning models. This is done by training a neural network to minimize the residuals of the governing equations in sample points. In this work, our flow is governed by the Navier-Stokes equations with some approximations. There are two main novelties in this paper. First, we design our model to assume that the flow is periodic in time, which is not feasible in conventional simulation methods. Second, we evaluate the benefit of resampling the function evaluation points during training, which has a near zero computational cost and has been verified to improve the final model, especially for small batch sizes. Finally, we discuss some limitations of the approximations used in the Navier-Stokes equations regarding the modeling of turbulence and how it interacts with PINNs.
△ Less
Submitted 20 December, 2022;
originally announced December 2022.
-
Optimal computational parameters for maximum accuracy and minimum cost of Arnoldi-based time-stepping methods for flow global stability analysis
Authors:
Marlon Sproesser Mathias,
Marcello Augusto Faraco de Medeiros
Abstract:
Global instability analysis of flows is often performed via time-stepping methods, based on the Arnoldi algorithm. When setting up these methods, several computational parameters must be chosen, which affect intrinsic errors of the procedure, such as the truncation errors, the discretization error of the flow solver, the error associated with the nonlinear terms of the Navier-Stokes equations and…
▽ More
Global instability analysis of flows is often performed via time-stepping methods, based on the Arnoldi algorithm. When setting up these methods, several computational parameters must be chosen, which affect intrinsic errors of the procedure, such as the truncation errors, the discretization error of the flow solver, the error associated with the nonlinear terms of the Navier-Stokes equations and the error associated with the limited size of the approximation of the Jacobian matrix. This paper develops theoretical equations for the estimation of optimal balance between accuracy and cost for each case. The 2D open cavity flow is used both for explaining the effect of the parameters on the accuracy and the cost of the solution, and for verifying the quality of the predictions. The equations demonstrate the impact of each parameter on the quality of the solution. For example, if higher order methods are used for approaching a Fréchet derivative in the procedure, it is shown that the solution deteriorates more rapidly for larger grids or less accurate flow solvers. On the other hand, lower order approximations are more sensitive to the initial disturbance magnitude. Nevertheless, for accurate flow solvers and moderate grid dimensions, first order Fréchet derivative approximation with optimal computational parameters can provide 5 decimal place accurate eigenvalues. It is further shown that optimal parameters based on accuracy tend to also lead to the most cost-effective solution. The predictive equations, guidelines and conclusions are general, and, in principle, applicable to any flow, including 3D ones.
△ Less
Submitted 8 November, 2022;
originally announced November 2022.
-
Enhancing Oceanic Variables Forecast in the Santos Channel by Estimating Model Error with Random Forests
Authors:
Felipe M. Moreno,
Caio F. D. Netto,
Marcel R. de Barros,
Jefferson F. Coelho,
Lucas P. de Freitas,
Marlon S. Mathias,
Luiz A. Schiaveto Neto,
Marcelo Dottori,
Fabio G. Cozman,
Anna H. R. Costa,
Edson S. Gomi,
Eduardo A. Tannuri
Abstract:
In this work we improve forecasting of Sea Surface Height (SSH) and current velocity (speed and direction) in oceanic scenarios. We do so by resorting to Random Forests so as to predict the error of a numerical forecasting system developed for the Santos Channel in Brazil. We have used the Santos Operational Forecasting System (SOFS) and data collected in situ between the years of 2019 and 2021. I…
▽ More
In this work we improve forecasting of Sea Surface Height (SSH) and current velocity (speed and direction) in oceanic scenarios. We do so by resorting to Random Forests so as to predict the error of a numerical forecasting system developed for the Santos Channel in Brazil. We have used the Santos Operational Forecasting System (SOFS) and data collected in situ between the years of 2019 and 2021. In previous studies we have applied similar methods for current velocity in the channel entrance, in this work we expand the application to improve the SHH forecast and include four other stations in the channel. We have obtained an average reduction of 11.9% in forecasting Root-Mean Square Error (RMSE) and 38.7% in bias with our approach. We also obtained an increase of Agreement (IOA) in 10 of the 14 combinations of forecasted variables and stations.
△ Less
Submitted 22 July, 2022;
originally announced August 2022.
-
Modeling Oceanic Variables with Dynamic Graph Neural Networks
Authors:
Caio F. D. Netto,
Marcel R. de Barros,
Jefferson F. Coelho,
Lucas P. de Freitas,
Felipe M. Moreno,
Marlon S. Mathias,
Marcelo Dottori,
Fábio G. Cozman,
Anna H. R. Costa,
Edson S. Gomi,
Eduardo A. Tannuri
Abstract:
Researchers typically resort to numerical methods to understand and predict ocean dynamics, a key task in mastering environmental phenomena. Such methods may not be suitable in scenarios where the topographic map is complex, knowledge about the underlying processes is incomplete, or the application is time critical. On the other hand, if ocean dynamics are observed, they can be exploited by recent…
▽ More
Researchers typically resort to numerical methods to understand and predict ocean dynamics, a key task in mastering environmental phenomena. Such methods may not be suitable in scenarios where the topographic map is complex, knowledge about the underlying processes is incomplete, or the application is time critical. On the other hand, if ocean dynamics are observed, they can be exploited by recent machine learning methods. In this paper we describe a data-driven method to predict environmental variables such as current velocity and sea surface height in the region of Santos-Sao Vicente-Bertioga Estuarine System in the southeastern coast of Brazil. Our model exploits both temporal and spatial inductive biases by joining state-of-the-art sequence models (LSTM and Transformers) and relational models (Graph Neural Networks) in an end-to-end framework that learns both the temporal features and the spatial relationship shared among observation sites. We compare our results with the Santos Operational Forecasting System (SOFS). Experiments show that better results are attained by our model, while maintaining flexibility and little domain knowledge dependency.
△ Less
Submitted 25 June, 2022;
originally announced June 2022.
-
The effect of Mach number on open cavity flows with thick or thin incoming boundary layers
Authors:
Marlon Sproesser Mathias,
Marcello Augusto Faraco de Medeiros
Abstract:
The Rossiter modes of an open cavity were studied using global linear analysis and nonlinear numerical simulations. The length over depth ratio was two and the Reynolds numbers based on cavity depth were close to 1000. The effect of Mach on such cavities was studied. The global analysis revealed that, in the Mach range 0.1 to 0.9, for thick boundary layers, only R1 and R2 modes could become unstab…
▽ More
The Rossiter modes of an open cavity were studied using global linear analysis and nonlinear numerical simulations. The length over depth ratio was two and the Reynolds numbers based on cavity depth were close to 1000. The effect of Mach on such cavities was studied. The global analysis revealed that, in the Mach range 0.1 to 0.9, for thick boundary layers, only R1 and R2 modes could become unstable, whereas for thin boundary layer up to R4 could be unstable. Compressibility was very destabilizing at low Mach. At moderate Mach the instability either saturated with Mach or had an irregular dependence. Analysis of the Rossiter mode eigenfunctions indicated that the acoustic feedback scaled to Ma^3, and explained the strong destabilizing effect of compressibility. The irregular dependence was associated with resonances between Rossiter modes and acoustic cavity modes. The analysis explained why the irregular Mach dependence occurred only for higher order Rossiter modes. In this parameter region three dimensional modes are stable or marginally unstable. Two-dimensional simulations were performed to evaluate how much of the nonlinear regime could be captured by the linear stability results. As the flow became more unstable, an increasing more complex final stage was reached. Yet the spectra present distinct tones that are close to linear predictions. However, a closer look reveals the dominant frequency in the saturated state corresponds to R1 while linear theory predicts R2 as the most unstable mode. This is associated with a nonlinear thickening of the mixing layer in the nonlinear regime. Moreover, the frequencies are much closer to Rossiter empirical predictions than the linear instability ones. Even at these conditions close to critical, the spectra are well described by the mode R1 and a cascade of nonlinearly generated harmonics, with no reminiscence of the linear instability.
△ Less
Submitted 27 May, 2022; v1 submitted 24 January, 2022;
originally announced January 2022.