Two exact first-order k-space formulations for low-rank viscoacoustic wave propagation on staggered grids

Wave propagation in the viscoacoustic media is physically dispersive and dissipated.Completely excluding the numerical dispersion error from the physical dispersion in the viscoacoustic wave simu-lation is indispensable to understanding the intrinsic property of the wave propagation in attenuated media for the petroleum exploration geophysics.In recent years,a viscoacoustic wave equation char-acterized by fractional Laplacian gains wide attention in geophysical community.However,the first-order form of the viscoacoustic wave equation,often solved by the conventional staggered-grid pseu-dospectral method,suffers from the numerical dispersion error in time due to the low-order finite-difference approximation.It is challenging to completely eliminate the error because the viscoacoustic wave equation contains two temporal derivatives,which stem from the time stepping and the amplitude attenuation terms,respectively.To tackle the issue,we derive two exact first-order k-space viscoacoustic formulations that can fully exclude the numerical error from the physical dispersion.For the homoge-neous case,two formulations agree with the viscoacoustic analytical solution very well and have the same efficiency.For the heterogeneous case,our second k-space formulation is more efficient than the first one because the second formulation significantly reduces the number of the wavenumber-space mixed-domain operators,which are the expensive part of the viscoacoustic k-space simulation.Nu-merical cases validate that the two first-order k-space formulations are effective and efficient alternatives to the current staggered-grid pseudospectral formulation for the viscoacoustic wave simulation.

一种模拟非均匀介质中弹性波传播新的k-space方法

传统的伪谱(PS)方法,采用傅里叶变换(FT)计算空间导数具有很高的精度,每个波长仅需要两个采样点,而时间导数采用有限差分(FD)近似因而精度较低.当采用大时间步长时,由于时空精度不平衡,PS法存在不稳定性问题.原始的k-space方法可以有效地克服这些问题但是却无法适用于非均匀介质.为了提高原始k-space方法模拟非均匀介质波动方程的精度,我们提出了一种新的k-space算子族.它是用非均匀介质的变速度代替原k-space算子中的常数补偿速度构造得到,引入低秩近似可以高效求解.我们将构造的新的k-space算子应用于耦合的二阶位移波动方程,而不是交错网格一阶速度应力波动方程,使模拟弹性波的计算存储量减少.我们从数学上证明了基于二阶波动方程的k-space方法与基于一阶波动方程的k-space方法是等价的.数值模拟实验表明,与传统的PS、交错网格PS和原始的k-space方法相比,我们的新方法可以在时间和空间步长较大的均匀和非均匀介质中,为弹性波的传播提供更精确的数值解.在保持稳定性和精度的同时,采用较大的时空采样间隔,可以大大降低数值模拟的计算成本.

Research of Sensitivity Encoding Reconstruction for MRI with Non-Cartesian K-Space Trajectories

The Sensitivity Encoding (SENSE) parallel reconstruction scheme for magnetic resonance imaging (MRI) is implemented with non-cartesian sampled k-space trajectories in this paper. SENSE has the special capability to reduce the scanning time for MRI experiments while maintaining the image resolution with under-sampling data sets. In this manner, it has become an increasingly popular technique for multiple MRI data acquisition and image reconstruction schemes. The gridding algorithm is also implemented with SENSE due to its ability in evaluating forward and adjoin operator with non-cartesian sampled data. In this paper, the sensitivity map profile, field map information and the spiral k-space data collected from an array of receiver coils are used to reconstruct unaliased images from under-sampled data. The performance of SENSE with real data set identifies the computational issues to be improved for researched.

基于K-means聚类和特征空间增强的噪声标签深度学习算法

深度学习中神经网络的性能依赖于高质量的样本,然而噪声标签会降低网络的分类准确率。为降低噪声标签对网络性能的影响,噪声标签学习算法被提出。该算法首先将训练样本集划分成干净样本集和噪声样本集,然后使用半监督学习算法对噪声样本集赋予伪标签。然而,错误的伪标签以及训练样本数量不足的问题仍然限制着噪声标签学习算法性能的提升。为解决上述问题,提出基于K-means聚类和特征空间增强的噪声标签深度学习算法。首先,该算法利用K-means聚类算法对干净样本集进行标签聚类,并根据噪声样本集与聚类中心的距离大小筛选出难以分类的噪声样本,以提高训练样本的质量;其次,使用mixup算法扩充干净样本集和噪声样本集,以增加训练样本的数量;最后,采用特征空间增强算法抑制mixup算法新生成的噪声样本,从而提高网络的分类准确率。并在CIFAR10、CIFAR100、MNIST和ANIMAL-10共4个数据集上试验验证了该算法的有效性。

6 W@80 K空间同轴脉管制冷机设计与实验研究

针对某空间红外载荷需求,基于Sage平台搭建了同轴脉管制冷机整机模型,根据仿真结果对冷头结构参数进行了设计计算。在此基础上,利用实验方法探究了回热器填充方式、惯性管组合方式、充气压力、输入功率、冷头方向对制冷机性能的影响,并与仿真结果比较,两者吻合度较高。研制成功的脉管制冷机整机质量为3.5 kg,输入功率150 W时,无负荷制冷温度为39.74 K,可获得7.23 W@80 K的冷量,整机相对卡诺效率13.26%,在空间用同轴脉管制冷机中处于先进水平。

基于改进的K-means聚类分区均匀化空间学习索引

传统空间索引的体量随数据量的增加而膨胀,查询效率较低。学习索引的体量不随数据量的增加而膨胀,同时避免了层级比较查询,性能优异。将学习索引应用于空间索引存在2个难点:一是选取合适的降维方法实现空间数据的排序;二是对降维后数据序列进行有效的简化分布计算,使其易于拟合。基于此,提出了一种网格混合聚类分区学习索引(grid-ml),用z曲线进行降维,用双层网格结构优化查询策略,用改进的K-means聚类算法进行数据分区,实现数据分布均匀化。对比实验发现,grid-ml构建速度快、存储空间小、查询效率高,较传统空间索引优势显著。

基于多尺度Harris角点检测的图像配准算法

针对现有多尺度Harris算子算法较复杂、运算量大、精确性一般的问题,提出一种高效简便算法。首先建立多尺度空间,令Harris算子在尺度空间提取特征点,用简化的32维SIFT特征向量描述特征。利用最近邻法匹配特征点;然后采用改进的相似三角形法筛选匹配点,再使用改进的K-means算法对特征点分组,使组内特征点聚集,组间特征点远离;最后应用改进的RANSAC算法在不同组中选取特征点求变换矩阵,避免了选取的特征点距离过近,算法陷入局部最优。实验验证了所提算法的性能。

SuccSite:Incorporating Amino Acid Composition and Informative k-spaced Amino Acid Pairs to Identify Protein Succinylation Sites

Protein succinylation is a biochemical reaction in which a succinyl group(-CO-CH2-CH2-CO-)is attached to the lysine residue of a protein molecule.Lysine succinylation plays important regulatory roles in living cells.However,studies in this field are limited by the difficulty in experimentally identifying the substrate site specificity of lysine succinylation.To facilitate this process,several tools have been proposed for the computational identification of succinylated lysine sites.In this study,we developed an approach to investigate the substrate specificity of lysine succinylated sites based on amino acid composition.Using experimentally verified lysine succinylated sites collected from public resources,the significant differences in position-specific amino acid composition between succinylated and non-succinylated sites were represented using the Two Sample Logo program.These findings enabled the adoption of an effective machine learning method,support vector machine,to train a predictive model with not only the amino acid composition,but also the composition of k-spaced amino acid pairs.After the selection of the best model using a ten-fold crossvalidation approach,the selected model significantly outperformed existing tools based on an independent dataset manually extracted from published research articles.Finally,the selected model was used to develop a web-based tool,SuccSite,to aid the study of protein succinylation.Two proteins were used as case studies on the website to demonstrate the effective prediction of succinylation sites.We will regularly update SuccSite by integrating more experimental datasets.SuccSite is freely accessible at http://csb.cse.yzu.edu.tw/SuccSite/.

基于聚类和随机搜索优化的核反应堆数字孪生参数反演模型

为了实现对核反应堆内置传感器的大量数据的高效存储、传输和分析,本文结合聚类算法与随机搜索优化的人工神经网络,对空间热离子反应堆的数字孪生系统搭建了一个参数反演模型,实现在热管失效工况下的堆芯温度数据的反演。使用20%热管失效工况下空间热离子反应堆堆芯4个区域的温度数据,通过K-means聚类与轮廓系数指标提取各区域的特征温度参数,通过随机配置优化的全连接人工神经网络(ANN)完成特征参数到其他参数的反演,并对反演效果进行验证。研究结果表明,运用该方法对燃料、发射极、接收极、冷却剂4个区域进行参数反演,温度反演值的相对误差均方根分别为0.55%、0.41%、0.19%、0.18%,其中用于反演的特征参数占总参数比例均不超过8%。因此本研究建立的参数反演模型能够获取特征参数的物理含义,并对空间热离子反应堆堆芯温度参数进行较高精度的反演。

BASIC PROPERTIES OF K-SPACES

Every complete metric linear space is a K-space.A non-completenormed space can be a K-space.And non-metrizable K-spaces exist.A series ofbasic results on complete metric linear spaces Can be generalized to K-spaces.Forexample,every bounded linear operator from a K-space into a locally convex spaceis sequentially continuous;every bounded semi-norm on a K-space is sequentiallycontinuous and,therefore,piecewise separable;a.C-sequeatial K-space is both bar-relled and bornological;the family of sequentially continuous linear functionals ona K-space is weak sequentially complete.

基于聚类和Apriori算法的苗族女上装色彩分析

苗族服饰元素是中国服饰文化的重要组成部分,其色彩搭配方式蕴含着极大的文化价值,展现了中国传统设计理念。以苗族女上装图像作为研究对象,对其色彩进行提取与分析。通过K-means聚类得到10个整体色、5个主色及10个辅色代表色,在HSV色彩空间下分析苗族服饰色彩分布,最后采用Apriori算法挖掘配色规则并建立配色网络。结果显示,苗族女上装以中低饱和、中低明度的红、黄、蓝色系为主,图案与主体色彩呈现出强烈对比。将配色关系应用到设计实践中,可促进苗族服饰色彩的传承与保护,为研究民族传统服饰色彩提供一定的借鉴和参考。

基于改进K均值聚类的光伏板缺陷检测方法

为了能够对光伏组件热斑部分准确地识别和提取,提出了一种基于HSV空间模型的改进K均值聚类图像处理方法。首先,将红外图像进行HSV空间转换和双边滤波处理,去除噪声并提高图像对比度;其次,使用高斯核函数估计实现图像灰度概率密度函数提取,并以此获取初始聚类中心;最后,利用先验知识对图像进行K均值聚类,提取和量化热斑缺陷。研究结果表明,该方法能够快速地检测定位热斑位置并统计出光伏板损坏程度,具有较高的精度以及较好的灵敏性和稳定性。

Top-k空间偏好查询方法研究

随着无线通信技术的发展和智能移动终端的广泛普及,基于位置的服务已经融入了人们生活的方方面面。其中,Top-k空间偏好查询在地理信息系统、城市建设规划、资源调度与分配、旅游规划等领域具有重要的意义。Top-k空间偏好查询是根据空间对象周围的特征对空间对象进行等级评价,并返回具有最高等级评价的k个空间对象。目前,对于Top-k空间偏好查询的研究主要集中在欧式空间和路网环境,本文对欧式空间和路网环境下的Top-k空间偏好查询方法进行分析和比较。

电能计量装置异常状态监测方法研究

电能计量装置计量过程中,受功率谱谐波震颤强度影响,其状态指标数据呈现高度混沌状态,导致监测性能较差。为此,提出基于改进聚类分析的电能计量装置异常状态监测方法。通过构建电能计量装置状态评估体系,选取具有评估价值的技术指标及其对应的参数指标,并将相关指标参数视为待分类对象,以获取电能计量装置运行状态的混沌样本数据。利用蚁群算法优化K-Means聚类方法的聚类指标数据,实现电能计量装置状态监测。试验结果表明:该方法应用后的电能计量装置异常状态检出率较高,并且预测值与实际值的拟合率、准确率均较高。该方法监测性能较好。

K-Spaces Property of Product Spaces

Let K be a class of spaces which are eigher a pseudo-open s-image of a metric space or a k-space having a compact-countable closed k-network. Let K′ be a class of spaces which are either a Fréchet space with a point-countable k-network or a point-G_δ k-space having a compact-countable k-network. In this paper, we obtain some sufficient and necessary conditions that the products of finitely or countably many spaces in the class K or K′ are a k-space. The main results are that Theorem A If X, Y ∈ K. Then X x Y is a k-space if and only if (X, Y) has the Tanaka’s condition. Theorem B The following are equivalent: (a) BF(w2) is false. (b) For each X, Y ∈ K′, X x Y is a k-space if and only if (X, Y) has the Tanaka’s condition.

New Approach to Synchronize General Relativity and Quantum Mechanics with Constant “K”-Resulting Dark Matter as a New Fundamental Force Particle

Planck scale plays a vital role in describing fundamental forces. Space time describes strength of fundamental force. In this paper, Einstein’s general relativity equation has been described in terms of contraction and expansion forces of space time. According to this, the space time with Planck diameter is a flat space time. This is the only diameter of space time that can be used as signal transformation in special relativity. This space time diameter defines the fundamental force which belongs to that space time. In quantum mechanics, this space time diameter is only the quantum of space which belongs to that particular fundamental force. Einstein’s general relativity equation and Planck parameters of quantum mechanics have been written in terms of equations containing a constant “K”, thus found a new equation for transformation of general relativity space time in to quantum space time. In this process of synchronization, there is a possibility of a new fundamental force between electromagnetic and gravitational forces with Planck length as its space time diameter. It is proposed that dark matter is that fundamental force carrying particle. By grand unification equation with space-time diameter, we found a coupling constant as per standard model “αs” for that fundamental force is 1.08 × 10-23. Its energy calculated as 113 MeV. A group of experimental scientists reported the energy of dark matter particle as 17 MeV. Thorough review may advance science further.

新能源场站电力二次系统安全预警方法研究

新能源场站电力二次系统设备较多。其安全预警性相关数据存在明显高维性、复杂性的特点,不利于电力二次系统安全预警的效率和精度计算,导致预警性能较差。为此,提出新能源场站电力二次系统安全预警方法。采用拉普拉斯映射方法降低新能源场站电力二次系统的电力数据维数,将高维数据源空间中的初始电力数据映射至低维子空间。分别计算在x、y、z方向上电力数据的行向量和对角权值,获取优化数据,为系统安全预警提供训练样本。从波动程度、时间序列变化趋势和能量随机分布特点三个角度,提取电力数据安全预警特征,并输入到K-means聚类算法中,以确定特征聚类中心、实现安全预警。试验结果表明:所提方法预警效率、预警精度较高。

应用自制克氏针外固定架治疗虎口挛缩

目的探讨应用自制克氏针外固定架治疗虎口挛缩的手术方法及疗效。方法2015年8月-2021年7月,使用自制克氏针外固定架治疗虎口挛缩20例,均有完整随访资料。采用臂丛神经阻滞麻醉,桡动脉鼻烟窝穿支皮瓣等方法修复虎口缺损后,在第1、2掌骨颈附近打入直径1.5 mm克氏针,并自制克氏针横梁,撑开虎口,弯曲克氏针端固定,依需要上下滑动横梁,调节虎口开大角度,固定4~6周后拆除。结果20例均获随访,随访时间4个月。拇指活动范围均明显增加,功能明显改善。结论使用自制克氏针外固定架配合皮瓣等方案治疗虎口挛缩,方法简单,经济实惠,治疗效果满意。

不同偏光系统下的岩石薄片图像联合分割

针对传统切换单偏光与正交偏光系统进行镜下岩石薄片分析和含量估计耗时耗力,且准确率较低的问题,在色度亮度空间设计一种岩石薄片图像联合分割方法:先将相同视域下的单偏光图像与正交偏光图像在色度亮度空间中分别进行分解,然后对色度图像进行双边滤波处理,抑制噪声和去除纹理,计算滤波后两个色度图像之间的差值,并依次使用K-Means聚类算法和分水岭分割算法得到最终分割结果。结果表明,使用单偏光岩石薄片图像与正交偏光岩石薄片图像联合分割较大提升了矿物颗粒计数准确率和分割结果的交并比。研究结果可为地质工作者鉴定岩石薄片提供指导。

基于相空间重构和PSO-K-means的球磨机负荷状态识别方法

针对球磨机振动信号具有强随机性、非平稳性和非线性等内在特性导致负荷状态难以识别的问题,提出一种基于相空间重构和PSO-K-means的球磨机负荷状态识别方法。首先,利用改进前后的自相关系数算法对Lorenz与Rossler两种混沌时间序列进行数值模拟,得出延迟时间和嵌入维数精准有效的计算方法;其次,验证出球磨机筒体振动信号具有混沌特性后对其时间序列进行相空间重构,恢复出等价的混沌吸引子;接着,针对3种不同负荷状态下的相空间吸引子进行特征提取,分析了关联维数特征量的变化规律;最后,将关联维数作为特征向量输入PSO-K-means聚类模型中对球磨机负荷状态进行分类与识别。结果表明,PSO-K-means聚类模型在负荷状态识别时有较高的精准性,欠负荷、正常负荷、过负荷下识别精度分别为94.2%、96.3%、94.8%。以上结果证实了该方法能够实现对球磨机负荷状态的有效识别。