Three dimensional (3D) microscopic distributions of dolomite and calcite in a limestone sample have been analyzed with a data-constrained modeling (DCM) technique using synchrotron radiation-based multi-energy X-ray c...Three dimensional (3D) microscopic distributions of dolomite and calcite in a limestone sample have been analyzed with a data-constrained modeling (DCM) technique using synchrotron radiation-based multi-energy X-ray computed tomography (CT) data as constraints. In order to optimize the experimental parameters, X-ray CT simulations and DCM analysis of a numerical phantom consisting of calcite (CaCO3) and dolomite (CaMg(CO3)2) have been used to investigate the effects on the predicted results in relation to noise, X-ray energy and sample-to-detector distance (SDD). The simulation results indicate that the optimal X-ray energies are 25 and 35 keVs, and the SDD is 10 mm. The high resolution 3D distributions of mineral phases of a natural limestone have been obtained. The results are useful for quantitative understanding of mineral, porosity, and physical property distributions in relation to oil and gas reservoirs hosted in carbonate rocks, which account for more than half of the world’s conventional hydrocarbon resources. The case studied is also instructive for the applicability of the DCM methods for other types of composite materials with modest atomic number contrasts between the mineral phases.展开更多
Big data analytics is often prohibitively costly. It is typically conducted by parallel processing with a cluster of machines, and is considered a privilege of big companies that can afford the resources. This positio...Big data analytics is often prohibitively costly. It is typically conducted by parallel processing with a cluster of machines, and is considered a privilege of big companies that can afford the resources. This position paper argues that big data analytics is accessible to small companies with constrained resources. As an evidence, we present BEAS, a framework for querying big relations with constrained resources, based on bounded evaluation and data-driven approximation.展开更多
提出了一种基于线性约束最小平方(Linear Constrained Least Square)方法的神经数据融合算法。LCLS方法用来最小化线性融合信息的能量，而神经网络算法则用来处理出现于LCLS方法中的样本协方差矩阵的不良条件和奇异性问题。此算法用软...提出了一种基于线性约束最小平方(Linear Constrained Least Square)方法的神经数据融合算法。LCLS方法用来最小化线性融合信息的能量，而神经网络算法则用来处理出现于LCLS方法中的样本协方差矩阵的不良条件和奇异性问题。此算法用软件和硬件都能实现。与已有的融合方法相比，文章提出的神经数据融合方法具有非偏倚的统计特性而且不需要关于噪声协方差的任何先验知识。将此方法应用于图像融合，结果显示这种方法能增强输出结果的质量。展开更多
文摘Three dimensional (3D) microscopic distributions of dolomite and calcite in a limestone sample have been analyzed with a data-constrained modeling (DCM) technique using synchrotron radiation-based multi-energy X-ray computed tomography (CT) data as constraints. In order to optimize the experimental parameters, X-ray CT simulations and DCM analysis of a numerical phantom consisting of calcite (CaCO3) and dolomite (CaMg(CO3)2) have been used to investigate the effects on the predicted results in relation to noise, X-ray energy and sample-to-detector distance (SDD). The simulation results indicate that the optimal X-ray energies are 25 and 35 keVs, and the SDD is 10 mm. The high resolution 3D distributions of mineral phases of a natural limestone have been obtained. The results are useful for quantitative understanding of mineral, porosity, and physical property distributions in relation to oil and gas reservoirs hosted in carbonate rocks, which account for more than half of the world’s conventional hydrocarbon resources. The case studied is also instructive for the applicability of the DCM methods for other types of composite materials with modest atomic number contrasts between the mineral phases.
基金湖南省科技公关计划( the Key Technologies R&D Program of Hunan Province China under Grant No.05GK2002)+2 种基金湖南省自然科学基金( the Natural Science Foundation of Hunan Province of China under Grant No.03JJY6023)长沙市科技公关计划( the Key Technologies R&D Program of Changsha City China under Grant No.K06070001- 12)
文摘Big data analytics is often prohibitively costly. It is typically conducted by parallel processing with a cluster of machines, and is considered a privilege of big companies that can afford the resources. This position paper argues that big data analytics is accessible to small companies with constrained resources. As an evidence, we present BEAS, a framework for querying big relations with constrained resources, based on bounded evaluation and data-driven approximation.
文摘提出了一种基于线性约束最小平方(Linear Constrained Least Square)方法的神经数据融合算法。LCLS方法用来最小化线性融合信息的能量，而神经网络算法则用来处理出现于LCLS方法中的样本协方差矩阵的不良条件和奇异性问题。此算法用软件和硬件都能实现。与已有的融合方法相比，文章提出的神经数据融合方法具有非偏倚的统计特性而且不需要关于噪声协方差的任何先验知识。将此方法应用于图像融合，结果显示这种方法能增强输出结果的质量。
