人工监督神经网络在煤层裂缝发育区识别中的应用

针对在地震属性应用中难以精细识别煤层裂缝发育区边界的难题,开展了基于人工监督神经网络技术的煤层裂缝发育区应用研究。从地震资料中获取倾角导向体以提取高质量的地震属性,以多属性为指导进行人工拾取样点,并基于多层感知器进行神经网络机器训练学习,建立裂缝的最优属性集,拓展整体数据后获得裂缝概率体,从而识别划分出煤层裂缝发育区。该技术在山西阳泉新元矿区进行了应用,煤层裂缝发育区的识别结果明显优于属性直接识别,但勘探区内暂无钻井资料,预测效果还有待进一步验证。     

A computer aided design method for car form and its application based on shape parameters

In the early design stage, automotive modeling should both meet the requirements of aesthetics and engineering. Therefore, a vehicle CAD (computer aided design) model that can be easily adjusted by feedbacks is necessary. Based on CE-Bézier surface, this paper presents a set of algorithms for parametric segmentation and fairing surface generation in a car model. This model is defined by a simplified automotive template and relevant control points, shape parameters and segmentation parameters, which can be modified to alter the car form efficiently. With this model and the corresponding adjustment method, more than fifty various vehicle models are established in this research according to different parameters. And two methods for calculating similarity index between car models are constructed, which are suitable for brand design trend analysis and modelling design decisionmaking.     

Self‐similarity analysis of vehicle driver's electrodermal activity

This paper characterizes stress levels via a self‐similarity analysis of the electrodermal activity (EDA) collected in a real‐world driving context. To characterize the EDA richness over scales, the fractional Brownian motion (FBM) process and its corresponding exponent H, estimated via a wavelet‐based approach, are used. Specifically, an automatic scale range selection is proposed in order to detect the linearity in a log scale diagram. The procedure is applied to the EDA signals, from the open database drivedb, originally captured on the foot and the hand of the drivers during a real‐world driving experiment, designed to evoke different levels of arousal and stress. The estimated Hurst exponent H offers a distinction in stress levels when driving in highway versus city, with a reference to restful state of minimal stress level. Specifically, the estimated H values tend to decrease when the driving environmental complexity increases. In addition, the estimated H values on the foot EDA signals allow a better characterization of the driving task than that of hand EDA. The self‐similarity analysis was applied to various physiological signals in literature but not to the EDA so far, a signal which was found to correlate most with human affect. The proposed analysis could be useful in real‐time monitoring of stress levels in urban driving spaces, among other applications.     

Discriminative graph convolution networks for hyperspectral image classification

Recently, the proposal of graph convolutional networks (GCN) has successfully implemented into hyperspectral image data representation and analysis. In spite of the great success, there are still several major challenges in hyperspectral image classification, including within-class diversity, and between-class similarity, which generally degenerate hyperspectral image classification performance. To address the problems, we propose a discriminative graph convolution networks (DGCN) for hyperspectral image classification. This method introduces the concepts of within-class scatter and between-class scatter, which respectively reflect the global geometric structure and discriminative information of the input space. The experimental results on the hyperspectral data sets show that the proposed method has good classification performance.     

基于孪生神经网络的时间序列相似性度量

在时间序列分类等数据挖掘工作中，不同数据集基于类别的相似性表现有明显不同，因此一个合理有效的相似性度量对数据挖掘非常关键。传统的欧氏距离、余弦距离和动态时间弯曲等方法仅针对数据自身进行相似度公式计算，忽略了不同数据集所包含的知识标注对于相似性度量的影响。为了解决这一问题，提出基于孪生神经网络（SNN）的时间序列相似性度量学习方法。该方法从样例标签的监督信息中学习数据之间的邻域关系，建立时间序列之间的高效距离度量。在UCR提供的时间序列数据集上进行的相似性度量和验证性分类实验的结果表明，与ED/DTW-1NN相比SNN在分类质量总体上有明显的提升。虽然基于动态时间弯曲（DTW）的1近邻（1NN）分类方法在部分数据上表现优于基于SNN的1NN分类方法，但在分类过程的相似度计算复杂度和速度上SNN优于DTW。可见所提方法能明显提高分类数据集相似性的度量效率，在高维、复杂的时间序列的数据分类上有不错的表现。     

粒子群属性聚类的位置隐私保护

针对基于位置服务中连续查询情况下，用户自身属性信息很容易被攻击者获取，并通过关联获得用户位置隐私的情况，提出了一种利用粒子群聚类加速相似属性用户寻找，并由相似属性匿名实现用户位置泛化的隐私保护方法。该方法利用位置隐私保护中常用的可信中心服务器，通过对发送到中心服务器中的查询信息进行粒子群属性聚类，在聚类的过程中加速相似属性用户的寻找过程，由相似属性用户完成位置泛化，以此实现位置隐私保护。实验结果证明，这种基于粒子群属性聚类的隐私保护方法具有高于同类算法的隐私保护能力，以及更快的计算处理速度。     

Comparative analysis of a Sisko nanofluid over a stretching cylinder through a porous medium

The present article examines the Sisko nanofluid flow and heat transfer through a porous medium due to a stretching cylinder using Buongiorno's model for nanofluids. Suitable similarity transformations are used to transform the governing boundary layer equations of fluid flow into nonlinear ordinary differential equations. The finite difference method is used to solve coupled nonlinear differential equations with MATLAB software. The impact of different parameters viz., the Sisko material parameter, porosity parameter, curvature parameter, thermophoresis parameter, and Brownian diffusion parameter on the velocity and temperature distribution are presented graphically. Moreover, the effect of the involved parameters on the heat transfer rate is also studied and presented through table values. It is noticed from the numerical values that the porosity parameter reduces the velocity while enhancing the temperature. The curvature parameter enhances the velocity throughout the fluid regime and reduces the temperature near the surface while enhancing the temperature far away from the surface. The study reveals that the thermophoresis and Brownian diffusion parameters that characterize the nanofluid flow reduce the wall heat transfer rate, while the curvature parameter enhances it. This investigation of wall heating/cooling has essential applications in solar porous water absorber systems, chemical engineering, metallurgy, material processing, and so forth.     

Test sample allocation method for testability verification test

An extended failure mode effect and criticality analysis (FMECA)-based sample allocation method for testability verification is presented in this study to deal with the poor representativeness of test sample sets and the randomness of the testability evaluation results caused by unreasonable selection of failure samples. First, the fault propagation intensity is introduced as part of the extended information of FMECA, and the sample allocation impact factors of component units and failure modes are determined under this framework. Then, the failure mode similarity and impact factor support are defined, and the game decision method for weighing the relationship between similarity and support is proposed to obtain the weight of failure mode impact factor. Finally, a two-step allocation framework of test samples is formulated to realize the sample allocation of component units and failure modes. This method is applied to the testability verification test of a launch control system. Results show that this method can obtain more representative test samples compared with the traditional sample allocation method while effectively reducing randomness of single testability evaluation result.     

Analysis and improvement of the influence of sample shape on Pu measurement results

ABSTRACT

The analysis method of Fast Neutron Multiplicity Counting (FNMC) plays an increasingly important role in the measurement of nuclear material properties. Based on the assumption of point model, fast neutron multiplicity measurement equation is derived which can be used to measure the mass of Pu sample. However, the deviation of the simulated measurement of 1 kg Pu sample reaches 16.6% and increases with mass. Because nonpoint source samples of different shapes do not fully satify the hypothesis. To correct this deviation, a set of fast neutron multiplicity counters was built by Geant4 to simulate and study the mass attribute of Pu samples.The cylindrical sources of different shapes and different masses were simulated, the self-multiplication factor and α coefficient were corrected.And the corresponding third-order polynomial fitting equation was obtained, the goodness of fit was greater than 0.970. In the same way, the spherical and spherical shell source samples in the mass range of 0–5 kg were analyzed, the corrected mass deviation of samples was less than 10% in this interval. The results show that the combination of the fast neutron multiplicity counter and parameter correction can accurately measure the sample mass attribute.     

基于链接相似度和作弊系数的Spam网页识别算法

Spam网页主要通过链接作弊手段达到提高搜索排名而获利的目的,根据链接作弊的特征,引入链接相似度和作弊系数两个指标来判定网页作弊的可能性。借鉴BadRank算法思想,从Spam网页种子集合通过迭代计算链接相似度和作弊系数,并根据与种子集合的链接指向关系设置权重,将待判定的网页进行度量。最后选取Anti-Trust Rank等算法作对比实验,结果验证了本文算法在准确率和适应性方面优于对比算法。