SimpleFWI

KO | EN

gitlite — search

#javascript #python #hacktoberfest #react #ai #typescript #llm #go #golang #android #machine-learning #rust #deep-learning #linux

(★ 55)

Simple Matlab code for testing optimization algorithms on seismic inverse problems.

SimpleFWI Latest Version Download

Download Latest Version (.zip)

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

SimpleFWI
=========

This code provides the basic building blocks to test optimization algorithms on seismic inverse problems.
The canonical seismic waveform inversion problem is given by

$$\min_{m} \sum_{i} \frac{1}{2}||P^TA(m)^{-1}q_i - d_i||_2^2 + \frac{\alpha}{2}||Lm||_2^2,$$

where $A(m)$ is the discretized Helmholtz operator $\omega^2 m + \nabla^2$ with absorbing boundary conditions, 
$P$ is the sampling operator, $q_i$ are the source vectors, $d_i$ are the observed data and $L$ is the discretized 
$\nabla$.

The main function is *misfit.m*, which takes as input the medium parameters $m$ and the data $d_i$ (and definitions of $P$, $q_i$, $\omega$ etc.) and returns the misfit value, its gradient and a function handle to compute the action of the Gauss-Newton hessian.

For an example, see */examples*, which uses a simple BB iteration to solve the optimization problem. 
To run these, you need to install the SPOT library: https://github.com/mpf/spot

Feel free to contact me with any questions, suggestions, etc.

Tristan van Leeuwen - T.vanLeeuwen@uu.nl

SimpleFWI

SimpleFWI Latest Version Download

Today's picks

Was this content helpful?

Recent Feedback

SimpleFWI

SimpleFWI Latest Version Download

Today's picks

Was this content helpful?

Recent Feedback

Download README