A. Kamkar Rouhani; M. Zakeri
Abstract
In order to obtain more accurate results from application of the method of artificial neural networks, instead of selection of the best network determined by trial and error process, we suitably combine the results of several networks that is called committee machine, to reduce the error, and thus, increasing ...
Read More
In order to obtain more accurate results from application of the method of artificial neural networks, instead of selection of the best network determined by trial and error process, we suitably combine the results of several networks that is called committee machine, to reduce the error, and thus, increasing the accuracy of the output results. In this research, ensemble combination of single artificial neural networks has been used in order to estimate the effective porosity of Kangan gas reservoir rock in South Pars hydrocarbon field. To achieve this goal, well logging data of 4 wells in the area at the depth interval corresponding to Kangan formation were used. Acoustic, density, gamma ray and neutron porosity well log data were assigned as the input of the networks while the effective porosity data were considered as the output of the networks. Back- propagation single neural networks having different structures were trained using regularization method and their results were assessed. Then, the networks with the best results, i.e. contained minimum mean of squares of errors in the test step, were selected for making ensemble combinations. To determine the weighting coefficients of the networks in the linear ensemble combinations, we applied three methods of simple averaging, Hashem’s optimal linear combination and non-analytical optimal linear combination employing genetic algorithm, and their results were compared. The best ensemble combination, in which we had the maximum reduction in mean of squares of errors of the test step compared to the best single neural network, was an optimal linear four-network combination obtained by using genetic algorithm optimization method. This best ensemble combination, compared to the best single neural network, reduced the mean of squares of errors in the training and test steps 3.6% and 11.2%, respectively.
M. Farzamian; A. Kamkar Rouhani; M. Ziaiie; H. A. Faraji Sabokbar; K. Seif panahi
Abstract
Chichakloo Lead and Zinc ore deposit is one of mineral potential areas, located in Lead and Zinc belt limit of Takab zone and 25 km far from Anguran mine. This ore deposit has been prospected and explored in different scales several times within the last few decades. The last exploration activity over ...
Read More
Chichakloo Lead and Zinc ore deposit is one of mineral potential areas, located in Lead and Zinc belt limit of Takab zone and 25 km far from Anguran mine. This ore deposit has been prospected and explored in different scales several times within the last few decades. The last exploration activity over the deposit is the detailed geophysical survey (using resistivity and induced polarization methods) and also geochemical survey for potential mapping of Lead and Zinc zones. In this paper, after modeling and interpretation of geophysical data and processing and interpretation of geochemical data, we have prepared suitable exploration maps in GIS environment. For this, we have taken a new fuzzy approach for exploration maps using trapezoidal membership function. Then, for integration of exploration fuzzy layers, we have used fuzzy operations. The results of investigation of the final integrated exploration map indicate lead and zinc zones having a fuzzy favorability of greater than 0.5 in southeast of the study area that is obtained from remarkable overlapping of geophysical and geochemical anomalies. The results of drilling boreholes in the area confirm the exploration results obtained in this research work.
Mahyar Yousefi; R. Gholami; A. Kamkarr-Ruhani; A. Moradzade
Abstract
In the systematic exploration plan for prospecting the mineral deposit, we can design an exploration algorithm using the modeling of known mineral occurrences. Such an algorithm is a key to recognize the area where is high probability of mineralization, reduce the risk of exploration and increases the ...
Read More
In the systematic exploration plan for prospecting the mineral deposit, we can design an exploration algorithm using the modeling of known mineral occurrences. Such an algorithm is a key to recognize the area where is high probability of mineralization, reduce the risk of exploration and increases the probability of exploration success. In this paper, we introduce an algorithm for optimizing mineral potential model and target generation in the exploration operation with focus on the gold exploration. In this way, after descriptive and conceptual modeling of gold deposit, all of the characteristics that can be used as an exploration criterion have been identified and assembled as a target model. Then, various data layers have been used to generate significant evidential maps. Then all of the evidential maps should be combined to generate mineral potential model (map) of the mineralization type sought. Recent map shows the probability location of gold mineralization as target area. Finally an algorithm has been introduced in which all of the exploration stage and methods have been identified base on priority.
M. Mohamadi Vizheh; A. Kamkar Rouhani
Abstract
Ground water, cavities, and isolate buried structures embedded at shallow depths are well detectable by resistivity and GPR methods because of distinct contrast in their electric and electromagnetic properties in comparison with their surrounding media. In this research work, 3 different profiles on ...
Read More
Ground water, cavities, and isolate buried structures embedded at shallow depths are well detectable by resistivity and GPR methods because of distinct contrast in their electric and electromagnetic properties in comparison with their surrounding media. In this research work, 3 different profiles on such targets have been chosen, and their responses have been investigated. Using both resistivity and GPR methods together, it has also been possible to investigate capabilities and limitations of the methods in practice. The results obtained from this research work indicate that the GPR method, in addition to its speed and simplicity in data acquisition, is very successful in detection of interfaces or boundaries between different media in which electromagnetic properties at the boundaries change rapidly. The resistivity surveys, which have been carried out using Wenner array in this study, indicate low resistivity of the media under investigation. The low resistivity of the subsurface media caused the depth of penetration of the GPR method to be low, and as a result, made it impossible to investigate the targets buried at depths greater than 2 meters. Unlike the GPR method, the resistivity method has not been very successful in detection of multiple targets with high resistivity contrasts. Lower resolution of the resistivity method in comparison with GPR method has caused this problem. In this study, considerable information has been obtained by selecting two different processing algorithms and applying them on a series of raw GPR dataset. The obtained information from the resistivities of the subsurface structures as a result of the resistivity surveys has made it possible to choose and apply these processing algorithms. This research work well indicates that high conductive areas in resistivity sections coincide with the areas in the GPR sections having intensive attenuation. This characteristic can be used well in the interpretation of the GPR sections. Finally the resistivity method can be introduced as a suitable supplementary geophysical method to the GPR method.