A non-iterative reconstruction method for an inverse problem modeled by a Stokes-Brinkmann equations

TL;DR

This paper introduces a non-iterative topology optimization method using topological gradients for reconstructing obstacles in a fluid governed by Stokes-Brinkmann equations, demonstrated through numerical examples.

Contribution

It presents a novel non-iterative approach employing topological gradients and level-set methods for obstacle reconstruction in fluid flow modeled by Stokes-Brinkmann equations.

Findings

Method effectively reconstructs obstacles from internal measurements.

The approach avoids truncation techniques used in previous methods.

Numerical examples confirm the method's efficiency and stability.

Abstract

This article is concerned with the reconstruction of obstacle $\O$ immersed in a fluid flowing in a bounded domain $\Omega$ in the two dimensional case. We assume that the fluid motion is governed by the Stokes-Brinkmann equations. We make an internal measurement and then have a least-square approach to locate the obstacle. The idea is to rewrite the reconstruction problem as a topology optimization problem. The existence and the stability of the optimization problem are demonstrated. We use here the concept of the topological gradient in order to determine the obstacle and it's rough location. The topological gradient is computed using a straightforward way based on a penalization technique without the truncation method used in the literature. The unknown obstacle is reconstructed using a level-set curve of the topological gradient. Finally, we make some numerical examples exploring the efficiency of the method.