Skip to content

Commit

Permalink
fixes
Browse files Browse the repository at this point in the history
  • Loading branch information
sunethwarna committed Feb 6, 2024
1 parent 890214e commit 5213f7a
Showing 1 changed file with 8 additions and 4 deletions.
Original file line number Diff line number Diff line change
@@ -1,14 +1,14 @@
---
title: Explicit Vertex Morphing
keywords:
keywords:
tags: [explicit, vertex morphing, filtering, optimization]
sidebar: optimization_application
summary:
summary:
---

## Introduction

Explicit Vertex morphing (here onwards called as vertex morphing) is a filtering method to filter out noise present in a gradient field or design field. Noisy gradient field is a common challenge in optimization problems which makes updated designs non-intuitive. They may also make the updated meshes based on updated design space with bad elements which eventually result in failure in primal and adjoint solution computation. Therefore it is crucial to apply a filtering technique to obtain a smooth gradient field. An example of a noisy field is shown in figure 1.
Explicit Vertex morphing (here onwards called as vertex morphing) is a discrete filtering method to regularize optimization problem. Non-smooth discrete gradient field is a common challenge in optimization problems which makes design updates noisy and non-applicable for numerical model, for instance, bad elements which eventually result in failure in primal and adjoint solution computation. Therefore it is crucial to apply a filtering technique to obtain a smooth design updates. An example of a noisy field is shown in figure 1.

<p align="center">
<img src="https://github.com/KratosMultiphysics/Documentation/blob/master/OptimizationApplication/Filtering/sensitivity_field_noise.png?raw=true" alt="Noisy design field"/>
Expand Down Expand Up @@ -132,8 +132,12 @@ Then for each vertex's (i.e. $$i^{th}$$) neigbhour vertices (i.e. $$j^{th}$$) th

<p align="center">$$ \left(\frac{df}{d\underline{s}}\right)_{i, morphed} = \frac{1}{\sum_{j=1}^N A_{ij}}\sum_{j=1}^{N} A_{ij}\left(\frac{df}{d\underline{s}}\right)_j$$</p>

## References
TODO:


## Source files
* [Doxygen](TODO) TODO
* [applications/OptimizationApplication/custom_utilities/filtering/filter_function.h](https://github.com/KratosMultiphysics/Kratos/blob/master/applications/OptimizationApplication/custom_utilities/filtering/filter_function.h)
* [applications/OptimizationApplication/custom_utilities/filtering/damping_function.h](https://github.com/KratosMultiphysics/Kratos/blob/master/applications/OptimizationApplication/custom_utilities/filtering/damping_function.h)
* [applications/OptimizationApplication/custom_utilities/filtering/explicit_filter.h](https://github.com/KratosMultiphysics/Kratos/blob/master/applications/OptimizationApplication/custom_utilities/filtering/explicit_filter.h)
* [applications/OptimizationApplication/custom_utilities/filtering/explicit_filter.h](https://github.com/KratosMultiphysics/Kratos/blob/master/applications/OptimizationApplication/custom_utilities/filtering/explicit_filter.h)

0 comments on commit 5213f7a

Please sign in to comment.