Interpretation of magnetic anomalies due to dipping dikes using neural network inversion

Abstract

A new approach is proposed for the interpretation of magnetic anomalies caused by dipping dikes. This approach is mainly based on modular neural network inversion for estimating the parameters of dipping dike model. Suitable network training examples and test data have been generated using forward models based on known true parameters. The training procedures adopt supervised learning routine using modular neural networks. The effect of random noise has been examined where the proposed technique showed stability and satisfactory results. The applicability of this technique has been tested on synthetic and field examples data. This technique is particularly applied to two field examples, namely magnetic anomaly over an outcropping quartz dike-like body in Karimnagar area, Andhra Pradesh, India, and Marcona magnetic anomaly, Marcona district, Peru. The results of using this technique showed good agreement with the measured field data compared with most conventional ones. Furthermore, neural networks proved to be efficient and flexible in the interpretation of magnetic anomaly of dipping dike.

Inversion of Vertical Electrical Sounding Data by Using General Regression Neural Network

“…Bescoby ve diğerleri [26] arkeolojik alanda toplanmış manyetik verilerin değerlendirilmesi için çok katmanlı yapay sinir ağlarını kullanmıştır. Al-Garni [27] ise manyetik verileri değerlendirirken modüler sinir ağlarını kullanmıştır. Modüler sinir ağları ayrıca El-Kalioubly ve Al-Garni [28] tarafından doğal potansiyel yönteminde uygulanmıştır.…”

Magnetic data interpretation for 2D dikes by the metaheuristic bat algorithm: sustainable development cases

“…On the other side, metaheuristic algorithms were developed to interpret the geomagnetic data, which rely on searching for global optimum solution that is more accurate and efficient than graphical and numerical methods 46 . Metaheuristic algorithms such as simulated annealing technique (SA) 47 , 48 , genetic algorithm (GA) 49 , particle swarm optimization (PSO) 50 , 51 , neural networks approach (NN) 22 , 52 , differential evolution algorithm (DE) 53 , and ant-colony optimization algorithm (ACO) 54 . These algorithms are popular among researchers because they are more adaptable and capable of dealing with a wide range of problems than traditional optimization techniques.…”

2D dipping dike magnetic data interpretation using a robust particle swarm optimization

“…Ram Babu and Atchuta Rao, 1991), Euler deconvolution method (Reid et al, 1990), Gauss-Newton method (Won, 1981), complex gradient method (Atchuta Rao et al, 1981), relation diagrams (Ram Babu et al, 1982), the gradient methods (Rao et al, 1973;Abdelrahman et al, 2007;Essa and Elhussein, 2016b), damped least-square ridge regression (Johnson, 1969), Spectral analysis methods (Bhattacharya, 1971;Sengupta and Das, 1975;Cassano and Rocca, 1975), modular neural network inversion (Al-Garni, 2015), an automated numerical method (Keating and 40 Pilkington, 1990), a new semi-automatic technique (Cooper, 2012), a non-linear constrained inversion technique (Beiki and Pedersen, 2012). However, the drawbacks of these methods are that they are highly subjective where they can lead to substantial errors in parameter estimations, rely upon trial and error till achieve the best fit between the measured and computed anomaly, require initial starting models which are close enough to the true solution, depends on the precision of separation of regional and residual magnetic anomalies from the measured magnetic anomaly,…”