Data preprocessing for parameter estimation. An application to a reactive bimolecular transport model

TL;DR

This paper introduces a data preprocessing methodology to optimize data collection intervals and simulate missing data, improving parameter estimation accuracy for reactive bimolecular transport models in porous media.

Contribution

It presents a novel preprocessing approach combining data interval selection and data simulation to enhance parameter estimation in reactive transport models.

Findings

The methodology improves parameter estimation accuracy.

Simulated data reduces experimental costs.

Application to porous media demonstrates effectiveness.

Abstract

In this work we are concerned with the inverse problem of the estimation of modeling parameters for a reactive bimolecular transport based on experimental data that is non-uniformly distributed along the interval where the process takes place. We proposed a methodology that can help to determine the intervals where most of the data should be taken in order to obtain a good estimation of the parameters. For the purpose of reducing the cost of laboratory experiments, we propose to simulate data where is needed and it is not available, a PreProcesing Data Fitting (PPDF).We applied this strategy on the estimation of parameters for an advection-diffusion-reaction problem in a porous media. Each step is explained in detail and simulation results are shown and compared with previous ones.