PASCAL - Pattern Analysis, Statistical Modelling and Computational Learning

Automatic wave-equation migration velocity inversion using multiobjective evolutionary algorithms
Vijay Pratap Singh, Bernard Duquet, Michel Léger and Marc Schoenauer
Geophysics Volume 73, VE61-VE73, 2008.


The presence of rugged topography, many folds and faults, complex velocity distribution, and irregular data coverage makes difficult to identify the over-thrust structures, either by seismic imaging or geological methods. We present an integrated automatic velocity extraction technique. In a previous work, we optimized a geological structure with respect to dip and fault location information. In this work, with the structure assumed to be known, we optimize the velocity field by using multi-objective evolutionary algorithms. Proposed technique is able to find a good velocity model., This is the first time, to the best of our knowledge, that multi-objective optimization is applied to a velocity optimization. These results, obtained on a known structure velocity field identification problem, show the power of the proposed representation. Direction of further research is the simultaneous identification of unknown structure and velocity field.

EPrint Type:Article
Project Keyword:Project Keyword UNSPECIFIED
Subjects:Learning/Statistics & Optimisation
ID Code:5121
Deposited By:Marc Schoenauer
Deposited On:11 April 2010