Gravity inversion of basement relief using imperialist competitive algorithm with hybrid techniques

Gravity profiles interpretation applying a metaheuristic particle optimization algorithm of mineralized bodies resembled by finite elements

“…Also, different optimization algorithms were applied to gravity modeling and interpretation such as genetic algorithm [ 32 ], bat algorithm [ 33 ], artificial bee colony (BCO) algorithm [ 34 ], cuckoo optimization algorithm [ 35 ], imperialist competitive algorithm [ 36 ], differential evolution algorithm [ 37 ], and Manta ray foraging optimization [ 38 ].…”

Gravity Inversion and Modeling Using Multi and Single objective Genetic Algorithms

“…In regional studies, residual gravity fields are generally ascribed to basement relief and Estimation of the depth of bedrock (or thickness of sedimentary basin) is a nonlinear inverse problem (Barbosa et al 1997;Nagihara and hall 2001;Chakravarthi and Sundararajan 2007;Tlas et al 2005;Abdeslsm 2017). Gravity survey data of a basin demonstrate that for the most part negative gravity anomaly values are associated with thick yet low-density valley-filling deposits overlaid on the bedrock (Schaefer 1983;Zhou 2012;Joolaei et al 2021). To investigate the geometry of the bedrock and the thickness of sediment on top of it in geophysics, local and global optimized methods are used (Jie and Tao 2015).…”

“…In this case, local search methods cannot be appropriate unless the primary model selection is close enough to the real response; it is, however, out of the control of geology structure complexity conditions (Snieder 1998;Tarantola 2005;Yuan et al 2008). Despite the execution simplicity and rapid convergence of local search methods, the probability of getting captured in the local minimum is high due to the reliance of these methods on the primary model and the existence of several optimized points in 2D or 3D modeling but the single-objective global optimization methods are biased toward the low misfit regions by means of some misfit criteria (Roy et al 2005;Redoloza and Li 2020;Joolaei et al 2021). The single-objective global optimization methods rarely need an initial model and in a lack of prior information they are more desirable than local optimization methods (Kaftan, 2017;Balkaya et al 2017).…”