基于混合网格Navier-Stokes方程的并行隐式计算方法研究

针对结构网格很难处理复杂外形和非结构网格无法计算具有边界层的粘性流动的缺点,发展了基于混合网格格点的隐式算法,成功地解决了在工程应用中难于处理的复杂外形粘性流场计算和效率问题。同时针对大规模的工程问题,发展了基于MPI通信技术的染色分层通讯并行计算方法。其中空间离散采用基于Roe格式发展的三阶迎风HLLEW(Harten-Lax-Van Leer-Einfeldt-Wada)或AUSM格式,湍流模型采用k??两方程湍流模型,时间推进考虑到LU-SGS并行等效较困难则采用基于DP-LUR(Data-Parallel Lower-Upper Relaxation)格式的隐式算法,计算CFL数可取到105量级,从2个到128个CPU的并行加速效率都保持在90%以上,大大提高了计算效率。算例对标模M6机翼模型流场进行计算,验证了方法的可靠性;然后对标模DLR-F6翼身组合体进行混合网格粘性与无粘计算结果进行比较,进一步验证混合网格方法;最后计算了DLR-WBNP外挂发动机翼身组合体模型,准确模拟了外挂和超临界机翼的相互干扰流动问题,采用4 CPU 16 CORE到24 CPU 96 CORE,2000步计算时间都不超...     

An Asynchronous Distributed Expectation Maximization Algorithm for Massive Data: The DEM Algorithm

The family of expectation--maximization (EM) algorithms provides a general approach to fitting flexible models for large and complex data. The expectation (E) step of EM-type algorithms is time-consuming in massive data applications because it requires multiple passes through the full data. We address this problem by proposing an asynchronous and distributed generalization of the EM called the distributed EM (DEM). Using DEM, existing EM-type algorithms are easily extended to massive data settings by exploiting the divide-and-conquer technique and widely available computing power, such as grid computing. The DEM algorithm reserves two groups of computing processes called workers and managers for performing the E step and the maximization step (M step), respectively. The samples are randomly partitioned into a large number of disjoint subsets and are stored on the worker processes. The E step of DEM algorithm is performed in parallel on all the workers, and every worker communicates its results to the managers at the end of local E step. The managers perform the M step after they have received results from a γ-fraction of the workers, where γ is a fixed constant in (0, 1]. The sequence of parameter estimates generated by the DEM algorithm retains the attractive properties of EM: convergence of the sequence of parameter estimates to a local mode and linear global rate of convergence. Across diverse simulations focused on linear mixed-effects models, the DEM algorithm is significantly faster than competing EM-type algorithms while having a similar accuracy. The DEM algorithm maintains its superior empirical performance on a movie ratings database consisting of 10 million ratings. Supplementary material for this article is available online.     

Message passing interface and multithreading hybrid for parallel molecular docking of large databases on petascale high performance computing machines

A mixed parallel scheme that combines message passing interface (MPI) and multithreading was implemented in the AutoDock Vina molecular docking program. The resulting program, named VinaLC, was tested on the petascale high performance computing (HPC) machines at Lawrence Livermore National Laboratory. To exploit the typical cluster‐type supercomputers, thousands of docking calculations were dispatched by the master process to run simultaneously on thousands of slave processes, where each docking calculation takes one slave process on one node, and within the node each docking calculation runs via multithreading on multiple CPU cores and shared memory. Input and output of the program and the data handling within the program were carefully designed to deal with large databases and ultimately achieve HPC on a large number of CPU cores. Parallel performance analysis of the VinaLC program shows that the code scales up to more than 15K CPUs with a very low overhead cost of 3.94%. One million flexible compound docking calculations took only 1.4 h to finish on about 15K CPUs. The docking accuracy of VinaLC has been validated against the DUD data set by the re‐docking of X‐ray ligands and an enrichment study, 64.4% of the top scoring poses have RMSD values under 2.0 Å. The program has been demonstrated to have good enrichment performance on 70% of the targets in the DUD data set. An analysis of the enrichment factors calculated at various percentages of the screening database indicates VinaLC has very good early recovery of actives. © 2013 Wiley Periodicals, Inc.     

Methods for wave equation prestack depth migration and numerical experiments

In this paper the methods of wave theory based prestack depth migration and their implementation are studied. Using the splitting of wave operator, the wavefield extrapolation equations are deduced and the numerical schemes are presented. The numerical tests for SEG/EAEG model with MPI are performed on the PC-cluster. The numerical re-     

SUN5500小型机MPI编程与蕴式并行计算的实现方法

基于SUN5500小型计算机并行开发环境，给出了消息传递模型和蕴式行模型的实现方法，通过实例分析了SUNMPI实际编程，并对选取不同模型有不同参数的运算时间进行了比较，结果表明，在SUN5500计算机上MPI模型和蕴式并行模型均能较大地提高运算速度，而且MPI在灵活性和并行程度方面更优。     

Study of correlation effects in high-multiplicity proton-proton reactions at 12 and 24 GeV/c

In this paper we study correlations in multiparticle final states from pp interactions at 12 and 24 GeV/c. In an attempt to distinguish true dynamical correlations from the consequences of kinematics together with P_T damping and the leading-particle effect, we compare the data with an independent-emission model which reproduces the single-particle spectra and also with a model that simulates a fragmentation mechanism. We investigate the forward-backward particle configurations and in particular the multiplicity imbalance and charge transfer, defining forward-backward by the largest rapidity gap as well as simply by c.m.s. hemispheres. We also study clustering by looking at distributions of the dispersions in longitudinal rapidity. From the comparison of the data with the models we find clear evidence for dynamical correlations of a sort one would expect from fragmentation-type mechanisms. We also find indications of non-fragmentation formation of neutral meson clusters.     

Inclusive production of Σ±(1385) in pp interactions at 12 and 24 GeV/c

The inclusive production of Σ^±(1385) is studied in pp interactions at 12 and 24 GeV/c. In this energy range, the inclusive cross sections for Σ⁺(1385) and Σ^?(1385) rise from 0.20 ± 0.03 mb to 0.28 ± 0.03 mb and from 0.07 ± 0.02 mb to 0.12 ± 0.02 mb. The decays of Σ^±(1385) account for ～20% of all observed Λ hyperons. The p_T² distributions are compatible with an exponential decrease and the slopes are in agreement with a common value of B ～ 3 (GeV/c)^?2. The longitudinal spectra are significantly different: Σ^?(1385) is mainly produced in the central region, whereas proton fragmentation contributes strongly to Σ⁺(1385) production.     

Multiplicities of secondary hadrons produced in vp and vp charged current interactions

In an experiment with the hydrogen bubble chamber BEBC at CERN multiplicities of hadrons produced in νp and $\text{v}\text{p}$ interactions have been investigated. Results are presented on the multiplicities of charged hadrons and neutral pions, forward and backward multiplicities of charged hadrons and correlations between forward and backward multiplicities. Comparisons are made with hadronic reactions and e⁺e^? annihilation. In the framework of the quark-parton model the data imply similar charged multiplicities for the fragments of a u- and a d-quark, and a larger multiplicities for the fragments of a uu- than for a ud-diquark. The correlation data suggest independent fragmentation of the quark and diquark for hadronic masses above ～ 7 GeV and local charge compensation within an event.     

A study of final states with a large transverse momentum particle from proton-proton collisions at 24 GeV/c

Events with a large transverse momentum (>1 GeV/c) pion or nucleon have been selected from the data of a high-statistics pp bubble chamber experiment (√s=6.84 GeV). Only events in which all secondary particles could be identified were used. One finds that fewer pions are produced in the azimuthal hemisphere containing the large transverse momentum particle than in the opposite hemisphere. An indication for coplanarity is found. Most pions associated with a large transverse momentum pion are emitted with small absolute c.m. rapidities, whereas those associated with large transverse momentum nucleons show some back-to-back structure. Various results of this investigation are similar to those obtained at the ISR. Most of the findings are compatible with predictions from an independent emission model.     

Quark confinement and topology of gauge theories

The phenomenon of quark confinement is known to be connected with the restoration of apparently broken gauge symmetry. In this paper we focus on a special mechanism which is responsible for such restoration. The major suggestion is that in the treatment of infrared problems certain classical field trajectories are of paramount importance, which trajectories connect seemingly degenerate vacua thereby eliminating the degeneracy. Initially, we demonstrate this mechanism in certain simple non-gauge models. In order of increasing difficulty, we next examine compact quantum electro-dynamics in 2 + 1 space-time dimensions- essentially a variant of unified models of the Georgi-Glashow type. In this case we prove that charge is confined and show that the force between two charges is independent of the distance between them. For small values of the fine structure constant this force is explicitly evaluated. Finally, we turn to the more realistic case of a 3 + 1 dimensional non-Abelian gauge theory, and analyse the contribution of a single pseudoparticle to the correlation functions. It is proposed that the quantum fluctuations of the pseudoparticle are inessential, and that the one-loop approximation is effectively correct even for the large scale pseudoparticles. The emergent conclusion then is that the renormalized Yang-Mills theory is reduced to the problem of evaluating the effects of those configurations which involve many pseudoparticles, due account being taken of the interactions between them. Some aspects of this last problem are also discussed.