Regularized focusing inversion for large-scale gravity data based on GPU parallel computing

Processing large-scale 3-D gravity data is an important topic in geophysics field. Many existing inversion methods lack the competence of processing massive data and practical application capacity. This study proposes the application of GPU parallel processing technology to the focusing inversion method, aiming at improving the inversion accuracy while speeding up calculation and reducing the memory consumption, thus obtaining the fast and reliable inversion results for large complex model. In this paper, equivalent storage of geometric trellis is used to calculate the sensitivity matrix, and the inversion is based on GPU parallel computing technology. The parallel computing program that is optimized by reducing data transfer, access restrictions and instruction restrictions as well as latency hiding greatly reduces the memory usage, speeds up the calculation, and makes the fast inversion of large models possible. By comparing and analyzing the computing speed of traditional single thread CPU method and CUDA-based GPU parallel technology, the excellent acceleration performance of GPU parallel computing is verified, which provides ideas for practical application of some theoretical inversion methods restricted by computing speed and computer memory. The model test verifies that the focusing inversion method can overcome the problem of severe skin effect and ambiguity of geological body boundary. Moreover, the increase of the model cells and inversion data can more clearly depict the boundary position of the abnormal body and delineate its specific shape.

High performance parallel computing of large eddy simulation of the flow in a curved duct with square cross section

Large eddy simulation(LES) cooperated with a high performance parallel computing method is applied to simulate the flow in a curved duct with square cross section in the paper. The method consists of parallel domain decomposition of grids, creation of virtual diagonal bordered matrix, assembling of boundary matrix, parallel LDL^T decomposition, parallel solving of Poisson Equation, parallel estimation of convergence and so on. The parallel computing method can solve the problems that are difficult to solve using traditional serial computing. Furthermore, existing microcomputers can be fully used to resolve some large-scale problems of complex turbulent flow.

PDBSCAN: Parallel DBSCAN for Large-Scale Clustering Applications

Density-based algorithm for discovering clusters in large spatial databases with noise(DBSCAN) is a classic kind of density-based spatial clustering algorithm and is widely applied in several aspects due to good performance in capturing arbitrary shapes and detecting outliers. However, in practice, datasets are always too massive to fit the serial DBSCAN. And a new parallel algorithm-Parallel DBSCAN(PDBSCAN) was proposed to solve the problem which DBSCAN faced. The proposed parallel algorithm bases on MapReduce mechanism. The usage of parallel mechanism in the algorithm focuses on region query and candidate queue processing which needed substantive computation resources. As a result, PDBSCAN is scalable for large-scale dataset clustering and is extremely suitable for applications in E-Commence, especially for recommendation.

A GPU-Based Parallel Algorithm for 2D Large Deformation Contact Problems Using the Finite Particle Method

Large deformation contact problems generally involve highly nonlinear behaviors,which are very time-consuming and may lead to convergence issues.The finite particle method(FPM)effectively separates pure deformation from total motion in large deformation problems.In addition,the decoupled procedures of the FPM make it suitable for parallel computing,which may provide an approach to solve time-consuming issues.In this study,a graphics processing unit(GPU)-based parallel algorithm is proposed for two-dimensional large deformation contact problems.The fundamentals of the FPM for planar solids are first briefly introduced,including the equations of motion of particles and the internal forces of quadrilateral elements.Subsequently,a linked-list data structure suitable for parallel processing is built,and parallel global and local search algorithms are presented for contact detection.The contact forces are then derived and directly exerted on particles.The proposed method is implemented with main solution procedures executed in parallel on a GPU.Two verification problems comprising large deformation frictional contacts are presented,and the accuracy of the proposed algorithm is validated.Furthermore,the algorithm’s performance is investigated via a large-scale contact problem,and the maximum speedups of total computational time and contact calculation reach 28.5 and 77.4,respectively,relative to commercial finite element software Abaqus/Explicit running on a single-core central processing unit(CPU).The contact calculation time percentage of the total calculation time is only 18%with the FPM,much smaller than that(50%)with Abaqus/Explicit,demonstrating the efficiency of the proposed method.

Parallel algorithm of solidification process simulation for large-sized system of liquid metal atoms

A parallel arithmetic program for the molecular dynamics (MD) simulation study of a large sized system consisting of 50 000100 000 atoms of liquid metals is reformed, based on the cascade arithmetic program used for the molecular dynamics simulation study of a small sized system consisting of 5001 000 atoms. The program is used to simulate the rapid solidification processes of liquid metal Al system. Some new results, such as larger clusters composed of more than 36 smaller clusters (icosahedra or defect icosahedra) obtained in the system of 50 000 atoms, however, the larger clusters can not be seen in the small sized system of 5001 000 atoms. On the other hand, the results from this simulation study would be more closed to the real situation of the system under consideration because the influence of boundary conditions is decreased remarkably. It can be expected that from the parallel algorithm combined with the higher performance super computer, the total number of atoms in simulation system can be enlarged again up to tens, even hundreds times in the near future.

Decentralized robust stabilization of discrete-time fuzzy large-scale systems with parametric uncertainties: a LMI method

Decentralized robust stabilization problem of discrete-time fuzzy large-scale systems with parametric uncertainties is considered. This uncertain fuzzy large-scale system consists of N interconnected T-S fuzzy subsystems, and the parametric uncertainties are unknown but norm-bounded. Based on Lyapunov stability theory and decentralized control theory of large-scale system, the design schema of decentralized parallel distributed compensation （DPDC） fuzzy controllers to ensure the asymptotic stability of the whole fuzzy large-scale system is proposed. The existence conditions for these controllers take the forms of LMIs. Finally a numerical simulation example is given to show the utility of the method proposed.

Shadow free texture acquisition of a large scale scene for city modeling

Texture acquisition of a large scale scene is one of the critical research areas in computer vision and can be used in other application areas such as computer graphics (CG), the intelligent transportation system (ITS) and the 3D geographic information system (GIS). Moreover, to acquire texture without noise (e.g., a shadow, an obstacle body) is vital for such work. Although obstacles can be removed by using 3D geometric data, shadow elimination is still a difficult problem and strongly required for the CG and ITS community, especially for city modeling and simulation purposes. In this paper, we propose an automatic multiple image fusion technique and an efficient and simple shadow removing technique to retrieve high quality texture images of an urban area. The image fusion can be efficiently achieved by epipolar plane image (EPI) analysis, and the shadow elimination can be successfully carried out by an illumination independent color clustering technique. The strength of this algorithm is that we can successfully fuse multiple images and eliminate shadows from the fused single image, especially in low dynamic range images, which have proven difficult using previous techniques.

A Class of Real-Time Parallel Combined Methods ofDigital Simulation for Large Systems

In this paper, a class of real-time parallel combined methods (RTPCM) of the digital simulation for a partitioned large system is presented. By means of combination of the parallelism across the system with the parallelism across the method, stiff and non-stiff subsystems are solved in parallel on parallel computer by a parallel Rosenbrock method and a parallel RK method, respectively. Their construction, convergence and numerical stability are discussed, and the digitalsimulation experiments are conducted.

电大山区地物环境中电波传播的电磁计算

在无人驾驶与无人机等新兴行业中,信号覆盖范围的要求较高,不仅仅在城市,在人迹罕至的山地、沙漠、森林中也需要无线信号的覆盖才能真正完成远程操控,这些地区更多需要考虑的是地势变化对电磁传播所带来的影响。计算电磁学中的一致性几何绕射理论方法是分析电大环境电磁问题的有效方法,使用计算电磁学的方法研究电磁波在山区地物环境中的传播规律。给出了一种建立不规则地形模型的新方法,可以通过数字高程的网格数据生成电磁算法可用的三次多项式曲面,使用多个立方曲面对不规则地形进行拼接,使用平均均方根误差验证模型数据的准确性。基于所得的地形数据,完成了并行的几何光学算法,并对区域电磁场的分布进行了仿真计算。选取了实际山区地物环境进行了实地测量,测量结果与仿真结果对比趋势一致,验证了该方法在非规则地形中电磁波传播分析中的有效性。考虑环境电磁计算的规模,建立了相应的并行策略,100核测试的并行效率可以保持在80%以上。

大模型融入云平台,信息化走向数智化

随着人工智能的快速发展,大模型已逐渐发挥其关键作用,成为驱动企业数字化转型的重要力量。主要探寻了大模型崛起引发的众多变革,回溯了人工智能的发展轨迹,分析了自建基础大模型所面临的挑战,强调引入基础大模型提供方需要与垂直行业企业联手,合作开发行业大模型。针对中小型企业在开发行业大模型时所面临的困扰,提出“模型即服务”及其相关工具链可以通过协助微调形成专有企业模型的观点;探讨了大模型兴起对云服务产生的变革;对云网协同、算网融合以及人工智能加速数智化发展进行了深入分析。在数字经济时代的背景下,进一步分析了人工智能如何赋能新一代信息技术创新。目前正处于国际经济发展调整期、信息化向数字化的演进期及网络技术创新的加速期,大宽带、大连接、大数据、大平台、大模型、大智能正在快速地融合发展。

隧道分岔形式对分流局部通风阻力特性的影响

目的:随着城市建设水平和强度提高,带匝道的分岔隧道日趋增多。针对分岔隧道通风设计难题,亟待获取隧道分岔形式对分流局部通风阻力特性的影响规律。方法:通过搭建同步满足阻力和压力相似的1/13缩尺分岔隧道通风试验平台,构建分岔隧道三维CFD模型,探究了分岔隧道不同分岔形式气流的流动特征和局部阻力系数,并进一步分析了分流结构与减速车道等隧道分岔形式特征对局部阻力系数的影响。结果:表明主线与匝道局部阻力系数随分流比增大呈现先减小后增大的趋势,且在分流比为0.4时达到最小。平行分岔隧道相比定向分岔隧道,主线局部阻力系数差别不大,而匝道局部阻力系数存在较明显的差异,分流比较小时,平行分岔隧道匝道局部阻力更大,分流比较大时则相反。增设减速车道,分流比较小时会减小匝道局部阻力系数,但会增加主线局部阻力系数,分流比较大时则相反。结论:研究结果可为分岔隧道的通风设计提供参考。

基于新课标的中小学生计算思维量表构建研究

对学生计算思维的培养已成为数字时代的核心议题,科学、精准的评价则是培养学生计算思维能力的基础。然而,针对义务教育阶段学生的计算思维测评,目前国内尚无依托相关课程标准制定的本土化的、权威的量表。为此,研究首次依托《义务教育信息科技课程标准(2022年版)》(以下简称“新课标”)中的计算思维定义及培养要求,共以12920名(N1=1029,N2=1458,N3=10433)小学生和初中生为研究样本,运用收敛混合方法构建一个面向我国教育实际的、经严格论证的中小学生计算思维量表。结果显示:经过两轮的收敛混合分析,修改后的计算思维量表共包含5个因子及15个题项,具有良好的内容效度;经过大样本实证检验分析,量表具有良好的信效度,且具有跨性别、年级和地区测量等值性,可以用来测量中小学生的计算思维水平。

基于真实世界数据的大型影像设备资源配置模型的构建与分析

目的构建基于真实世界数据(Real World Data,RWD)的计算机断层扫描(Computed Tomography,CT)、磁共振成像(Magnetic Resonance Imaging,MRI)设备的资源配置模型,以优化公立医院大型影像设备的配置问题。方法选取我院2018年1月至2023年12月医学影像科CT、MRI设备作为研究对象,使用年投资收益率指标分级评价设备的经济效益,然后采用Pearson相关系数法分析医院门诊人次、出院人次、CT/MRI设备检查量与时间的相关性,最后构建基于数据拟合与回归分析的配置预测模型并进行分析评价。结果CT和MRI设备检查量与住院人次的Pearson相关系数r分别为0.557和0.851(P<0.01),住院人次、CT和MRI设备检查量与月份的r分别为0.415、0.559和0.731(P<0.01),相关性强。通过数据拟合与回归分析得到三次配置拟合模型更符合我院实际情况,在未来3年内至少需要增配1台CT和2台MRI设备以顺应医院高速发展需求。结论构建基于RWD的CT、MRI设备的资源配置模型,有效地优化了大型影像设备的配置规划,为公立医院配置大型影像设备提供决策依据。

MapReduce模型在大规模数据并行挖掘中的应用

MapReduce并行编程模型通过定义良好的接口和运行支持库,能够自动并行执行大规模计算任务,隐藏底层实现细节,降低并行编程的难度。系统阐述MapReduce的基本工作原理及其工作流程,以TeraSort算法为例,针对其存在的问题,提出动态数据分区和数据压缩等优化建议。研究成果表明,优化后的TeraSort算法能够显著缩短数据处理时间,优化系统的吞吐量,并改善资源分配的均衡性。

超大规模数据处理中并行计算技术的应用研究

随着人工智能和大数据时代的到来,超大规模数据处理成了一个重要的研究领域。该文主要探讨并行计算技术在超大规模数据处理中的应用,首先详细阐述并行计算和超大规模数据处理的基本理论与概念,特别是并行计算的编程模型与工具,最后通过分析并行计算在搜索引擎、气象预报和金融分析等中的实际案例,阐述并行计算技术在超大规模数据处理中的实际应用。

大规模协同虚拟环境下并行层次兴趣匹配算法

大规模协同虚拟环境运行过程中易产生大量冗余通信数据,影响虚拟现实技术性能的发挥。为了提升虚拟环境运行效率,提出并行层次兴趣匹配算法。创建协同虚拟环境概念模型,分解处理订阅区域信息,引入树结构理论搭建兴趣管理树,将订阅区域映射到兴趣管理树节点上,形成层次化结构,结合LSQR并行算法,实现更新区域与订阅区域的快速匹配。实验结果表明,提出算法获得的兴趣匹配消耗时间最小值为0.36 s,兴趣匹配成功指数最大值为0.94,充分证实了提出算法兴趣匹配性能较好。

人工智能与生成:计算机介入产品设计

人工智能技术的更新与迭代“有效加速”,其已开始介入产品设计实践生产。基于人工智能的生成原理与方法,本质上是计算机向操作者深度学习的过程。针对产品设计领域,计算机开始由“辅助”设计走向“介入”设计,计算机智能平台将审美转变为算法。从概念设计环节影响产品设计流程,形成设计师和计算机多元化的主体创作方式。产品设计师与人工智能技术“超级对齐”,合力推动产品设计过程高效开展。

智算中心高性能网络流量调度技术研究及实践

AI大模型训练、高性能存储等业务应用场景提出了海量规模的计算需求,与传统数据中心业务相比,在流量模型和网络需求方面有着显著区别,驱使传统的数据中心网络向智算中心和无损网络转型。从智算中心和无损网络的发展背景入手,分析了当前智算中心网络存在的问题,探索了智算中心网络流量调度的关键技术,并进行了流量调度平台的研发实践,为智算中心网络发展和应用提供思路。

面向大规模异构计算平台的MiniGo高效训练方法

提出一种适用于大规模异构计算平台训练MiniGo智能体的高效多级并行训练方法,包括节点间任务级并行、中央处理器-数字信号处理器(central processing unit-digital signal processor, CPU-DSP)异构并行、DSP核内并行。实现了高效的输入/输出部署,消除网络通信瓶颈。提出了面向CPU-DSP共享内存结构的异构计算内存管理,减少异构设备间的数据搬运。实现了共享内存编程优化,并利用DSP实现密集卷积计算算子加速优化。结果表明,与16核CPU计算相比,单核DSP算子加速最大加速比达16.44;该方法实现计算节点规模从1 067扩展至4 139,得到达到给定终止条件所需时间从43.02 h降至16.05 h,可扩展效率为69.1%。评估表明,该方法能够实现MiniGo在大规模异构计算平台的高效并行训练。

面向工业大模型的算力网络架构与关键技术

随着第四次工业革命的推动,工业生产逐渐迈入了数字化和智能化的时代。在这一时代背景下,工业大模型作为推动工业创新的核心引擎,扮演着越来越重要的角色。同时,工业大模型的广泛应用也给算力网络提出了更为复杂和严峻的实时性需求。本文深入研究了工业大模型在数字化转型中的关键问题,着眼于其对算力网络的需求,分析了工业大模型在实时性方面的复杂需求,提出了面向工业大模型的算力网络架构,并对算力网络的关键技术进行了介绍,为工业大模型的高效运行提供了技术支持,为工业数字化、智能化的快速发展提供了助力。