Smooth 3D Modeling of Gravity Data Using Particle Swarm Optimization
New Trends in Geophysics and Engineering 2018, İstanbul, Turkey, 7 - 09 November 2018, pp.1-4
- Publication Type: Conference Paper / Full Text
- City: İstanbul
- Country: Turkey
- Page Numbers: pp.1-4
- Istanbul University Affiliated: Yes
Abstract
In this study, the ability of the Global Optimization methods to solve geophysical modeling problems with large parameter counts is investigated. For this purpose, a Particle Swarm Optimization algorithm is developed and implemented to realize 3D modeling of gravity data.
The developed algorithm uses several approaches to cope with large amount of model parameters, generally encountered during 3D modeling. The algorithm employs a starting model and starts the search for the smoothest possible model iteratively. The algorithm decreases the smoothness as the iterations increase. In order to prevent artefacts and to avoid problems arising from the randomness of the method the minimum gradient support is added as penalty to the function to be optimized.
The developed algorithm is tested on a synthetic data using a model mesh with 32.000 cells. The algorithm is found to be able to model both datasets with the given model meshes.