Influencia do Comprimento do passo na Inversão da Forma de Onda (FWI)
Author(s) -
Rafael Abreu de Cristo,
Milton Jose Porsani
Publication year - 2016
Publication title -
proceedings of the 4 simpósio brasileiro de geofísica
Language(s) - English
Resource type - Conference proceedings
DOI - 10.22564/7simbgf2016.160
Subject(s) - gradient descent , inversion (geology) , maxima and minima , algorithm , geology , computer science , mathematics , seismology , mathematical analysis , artificial neural network , artificial intelligence , tectonics
The waveform inversion (FWI) is a technique to obtaining seismic velocity field from the fit between the observed and calculated data. In the oil industry it has been widely used due the fact improve lateral resolution of seismic section, it can be inferred more accurately reservoirs before undetected. Due to the high nonlinearity of the problem, and high computational cost, this problem has become challenging, in order to find a best fit between observed and calculated data in a shortest possible time. Therefore, for each current model adopts is a search towards the global minimum and stride length toward this minimum. This study aims to estimate the best step length between two step lengths calculation methods tested for this objective function is as small as possible and that it occurs in a shorter time as possible. In the calculation of the search direction, two methods can be used: steepest-descent or LBFGS, among others. This work used the steepest-descent method.
Accelerating Research
Robert Robinson Avenue,
Oxford Science Park, Oxford
OX4 4GP, United Kingdom
Address
John Eccles HouseRobert Robinson Avenue,
Oxford Science Park, Oxford
OX4 4GP, United Kingdom