New Approaches in Regularization of 2D Modelling of Potential Data

عنوان دوره: نوزدهمین کنفرانس ژئوفیزیک ایران
کد مقاله : 1622-NIGS
نویسندگان
1گروه فیزیک زمین، موسسه ژئوفیزیک، دانشگاه تهران، تهران، ایران
2عضو هیات علمی دانشگاه تهران
چکیده
In this paper we presented a novel method for inversion of geophysical data while the regularization term is implied in a meta-heuristic algorithm. Due to the integration of the regularization term in the body of the stochastic algorithm, some of the main challenges such as defining a weighting matrix and finding the suitable regularization factor are resolved. This paper elaborates some of the main ideas of indirect regularization of geophysical data for genetic algorithm and uses a noisy synthetic model to validate the stability and efficiency of these techniques. All computations have been done in the GNU octave 5.1.0 environment.
کلیدواژه ها
 
Title
New Approaches in Regularization of 2D Modelling of Potential Data
Authors
Mahmoud Reshadati, Seyed-Hani Motavalli-Anbaran
Abstract
In this paper we presented a novel method for inversion of geophysical data while the regularization term is implied in a meta-heuristic algorithm. Due to the integration of the regularization term in the body of the stochastic algorithm, some of the main challenges such as defining a weighting matrix and finding the suitable regularization factor are resolved. This paper elaborates some of the main ideas of indirect regularization of geophysical data for genetic algorithm and uses a noisy synthetic model to validate the stability and efficiency of these techniques. All computations have been done in the GNU octave 5.1.0 environment.
Keywords
Regularization, Gravity data modelling, stochastic, Metaheuristic algorithms