21 June 2004 in Environmetrics
31 March 2023 in Applied Stochastic Models in Business and Industry
6 December 2025 in Applied Mathematical Modelling
1 December 2023 in Ciência e Natura
In this work, the inverse problem of parameter estimation in the advection-dispersion-reaction equation, modelling the pollutant dispersion in a river, is studied with a Neural Network approach. In the direct problem, the dispersion, velocity and reaction parameters are known and then the initial and boundary value problem is solved by classical numerical methods, where it is used as input dataset for the inverse problem and formulation. In the inverse problem, we know the dispersion and the velocity parameters and also the information about the pollutant concentration from the synthetic experimental data, and then the aim is to estimate the reaction parameter in the advection-dispersion-reaction equation. This inverse problem is solved by an usual Artificial Neural Network (ANN) and by a Physics-Informed Neural Network (PINN), which is a special type of neural networks that includes in its formulation the physical laws that describe the phenomena involved. Numerical experiments with both the ANN and PINN are presented, demonstrating the feasibility of the approach considered.
4 November 2024 in Ciência e Natura
The present work aims to formulate and solve the inverse problem of structural damage identification using Bayesian Inference. In the solution of the direct problem, the Finite Element Method (FEM) is considered. The modeling of the damage field is performed through the cohesion parameter, which continuously describes the integrity of the structure. The damage identification problem is formulated as an inverse parameter estimation problem, where the posterior probability distribution of the cohesion parameters is sampled using the Adaptive Markov Chain Monte Carlo method and a Spike-Slab prior, adopting a novel hierarchical modeling approach for the inverse problem and an appropriate prior distribution that naturally models the available information about the parameters of interest.
14 July 2023 in Defect and Diffusion Forum
ANAntônio José da Silva Neto
Antônio José da Silva Neto
In this work, we propose a stochastic wind field based on the Bayesian dynamic linear model to account for the wind flow field in the transient advection-diffusion partial differential equation (PDE). The resulting dispersion model accounts for the time variation in the wind field and meteorological variables, allowing the simulation of a transient regime. The main advantage of using such a wind field model over a Fourier series to fit wind time series is its potential to make predictions. In addition, a suitable methodology is necessary to solve the resulting dispersion model. In this work, we use a finite element formulation appropriate to solve transient advection-diffusion PDEs. We verify the accuracy of the proposed methodology by reproducing a case study considering a field tracer experiment. The model evaluation against experimental data shows the good performance of the proposed dispersion model.