OpenMP在水动力数学模型并行计算中的应用

采用PGI Fortran7.1-2(Portland Group)的OpenMP技术对二维水动力数学模型进行了并行优化试验。并行后,数学模型运行时间明显减少,在对同一算例的水流计算模拟中,串行和并行所需运行时间分别为5 336.781 s和3 454.296 s,平均加速比为1.56,平均并行计算节省时间36%,明显提高了水动力学数学模型的运算速度。     

非差非组合与组合模型定位性能对比及并行设计

分别采用非差无电离层组合模型与非差非组合模型进行批量静态定位解算,对比两者的定位精度以及ZPD估计精度。设计并实现两种模型的并行解算方法,提高非差模型的计算效率。大网数据实验结果表明,两种模型定位精度基本一致,非差非组合ENU的3个方向平均偏差为(4.2,2.8,6.6)mm,组合模型的平均偏差为(4.0,2.8,6.3)mm,但非差非组合模型的解算更加耗时,约是组合模型计算时间的1.4倍。多核环境下,两种模型的并行解算效率均比传统串行模型得到了提高。结果表明,双核并行和四核并行的计算效率比单核串行计算均分别提高了40%和60%以上。     

POM海洋模式的并行算法

POM模式目前尚无正式发布的并行版本。通过对POM串行程序的数据流向分析,讨论了POM模式并行化所涉及的关键算法和主要技术问题;并基于消息传递接口(MPI),研发了POM模式的并行版本。测试结果表明,POM并行软件效率较高,达到了业务化要求,业已应用于国家海洋信息中心的再分析业务化系统中。     

MPI环境下基于边界校正方法的海洋模型的并行化

基于区域分解技术,设计了一种针对海洋模型的边界校正方法,该方法不需要对原有海洋模型进行较大改动,只需要利用消息传递接口(message passing interface,MPI)编写子区域间的数据传输模块,即可实现原有海洋模型的并行化。相对于重新开发并行系统,该方法能够较快把原有模型的串行解决方案转化为并行解决方案,并且并行化后的海洋模型能够获得较高的加速比。     

潮流场数学方程交错分区并行计算方法研究

文章对潮流场数学方程进行可并行性方法研究。分析了两种可并行实现的方法，并对两种可并行化的方法可行性进行分析。最后采用虚拟区域分解法对问题进行了实现，并利用串行计算的结果对并行计算的结果进行了验证。     

基于MPI的LAGFD-WAM海浪数值模式并行算法研究

在LAGFD-WAM海浪串行数值模式基础上,利用MPI信息传递机制实现其并行化。通过对模拟区域合理划分,对数据采取分块加载,实现了各个节点的负载平衡;通过对算法的改进实现了粗粒度计算,大幅度减少了通信量,从而提高了程序的执行效率。对串行计算和并行计算的效率比较表明,本文建立的方法能够得到较高的加速比。对全球海浪模式,加速比和CPU数目大致呈线性关系。对高分辨率的区域海浪模式,在128 CPU条件下加速比可以达到91.9。     

二维风暴潮模式并行计算研究

通过对串行程序的分析、探讨,提出了基于区域分解的二维风暴潮模式并行算法;通过生成派生数据类型的方式,解决了二维数组子块传送和不同类型数组对应并行化读写的问题;通过子程序全局通信的方式,解决了边界处理的复杂性,提高了模式的可扩展性.并在选取不同的网格规模、不同的进程数的条件下进行了测试,得到了较为理想的结果.     

一维Burgers方程的一类交替分段并行算法

研究并行算法解决应用并行计算机完成规模尽可能大的偏微分方程的数值求解问题。利用Hopf-Cole变换,将一维非线性Burgers方程转化为线性扩散方程,基于第二类Saul’yev型非对称格式和Crank-Nicolson格式对扩散方程进行差分离散,建立解Burgers方程的交替分段并行差分格式,并讨论该方法的稳定性,给出了数值算例。此算法把剖分节点分成若干组,在每组上构造能够独立求解的差分方程,因此具有并行本性,适合在高性能多处理器的并行计算机上使用。数值试验的结果表明此方法是有效的,且有较高的精度。     

传感器网络高频数据采集算法

随着传感器网络技术的发展,传感器被广泛地应用到各个领域。由于传统的传感器数据采集采用串行采集算法,采集速率较低,严重影响了整个传感器网络的效率。针对此问题,本文提出了一种并行的高频数据采集算法,批量进行数据采集,并改进了设备响应等待时间机制。通过大量的实验数据进行对比分析及可视化处理,并进行验证分析,最终结果表明该算法在保证数据采集正确率的同时,大幅提高了数据采集的速率,并且数据采集的对象越多,提高的越明显。     

一个基于GPU并行加速的海啸数值模型

地震海啸通常发生在大洋板块向陆地板块俯冲的区域,距离震源最近的国家和地区往往在震后5～20 min之内就会遭受到海啸袭击。因此,及时的海啸预警和准确的海啸预报结果对于民众和决策者都至关重要。为了提升海啸预警效率,缩短海啸预报时间,本研究对COMCOT海啸数值模型进行了基于图形计算单元GPU的二次并行开发。将原模型中海啸传播计算模块通过CUDA＿C语言编写内核函数整体移植到GPU上并行加速,CPU负责模型其他代码的执行。为了减少CPU和GPU之间的数据通信,将吸收边界和变量更新函数一并改写。仅在需要输出的时间节点,GPU向CPU传递结果,其他时间步长,CPU和GPU之间只有指令和少量参数传输,基本可视为零耗时。基于GPU并行加速的COMCOT较串行版本效率提升超过67倍,加速性能显著优于基于CPU共享内存的OpenMP并行版本。交叉使用常水深和真实地形,采用均匀滑移海啸源和有限元海啸源对模型的计算结果进行了较为全面的分析检验,相对误差最大不超过1%,为大范围的越洋海啸实时计算提供了有力工具。     

Evaluation of a high-speed serial link for marine data acquisition

Evaluation of the Gazelle Microcircuits, Inc., HOT ROD high-speed (400 Mbit/sec) gallium arsenide serial link indicates that the device has tremendous potential as a multiplexer/demultiplexer designed to telemeter data through fiber-optic media. The TTL compatibility of all data I-O lines, transparent link operation, and synchronous mode of operation create a near-ideal environment for interfacing the HOT ROD chipset into external systems. The 40 bit parallel bus architecture of the chipset allows for great flexibility in system-level design, including the capability of simultaneous transmission of parallel and serial data. The chipset appears to be capable of extremely low bit error rates.     

An unstructured-grid, finite-volume, nonhydrostatic, parallel coastal ocean simulator

A finite-volume formulation is presented that solves the three-dimensional, nonhydrostatic Navier–Stokes equations with the Boussinesq approximation on an unstructured, staggered, z-level grid, with the goal of simulating nonhydrostatic processes in the coastal ocean with grid resolutions of tens of meters. In particular, the code has been developed to simulate the nonlinear, nonhydrostatic internal wave field in the littoral ocean. The method is based on the formulation developed by Casulli, in that the free-surface and vertical diffusion are semi-implicit, thereby removing stability limitations associated with the surface gravity wave and vertical diffusion terms. The remaining terms in the momentum equations are discretized explicitly with the second-order Adams–Bashforth method, while the pressure-correction method is employed for the nonhydrostatic pressure in order to achieve overall second-order temporal accuracy. Advection of momentum is accomplished with an Eulerian discretization which conserves momentum in cells that do not contain the free surface, and scalar advection is discretized in a way that ensures consistency with continuity, thereby ensuring local and global mass conservation using a velocity field that conserves volume on a local and global basis. The nonhydrostatic pressure field is solved efficiently using a block-Jacobi preconditioner, and while stability is limited by the internal gravity wave speed and vertical advection of momentum, applications requiring relatively small time steps due to accuracy or stability constraints are run efficiently on parallel computers, since the present formulation is written entirely with the message-passing interface (MPI). The ParMETIS libraries are employed in order to achieve a load-balanced parallel partitioning that minimizes interprocessor communication, and the grid is reordered to optimize per-processor performance by limiting cache misses while accessing arrays in memory. Test cases demonstrate the ability of the code to efficiently and accurately compute the nonhydrostatic lock exchange and internal waves in idealized as well as real domains, and we evaluate the parallel efficiency of the code using up to 32 processors.     

Oblique wave attack on rubble mound breakwaters

Stability formulae for armour layers of rubble mound breakwaters are usually being applied assuming perpendicular wave attack. Often the effects of oblique waves are neglected. This is however a conservative assumption since the stability of armour slopes generally increases for oblique waves. New wave basin tests have been performed to assess the effects of oblique waves on the stability of rock slopes and the stability of cube armoured rubble mound breakwaters. The physical model tests were focussed on wave directions between perpendicular (0°) and parallel (90°). The test programme included tests with long-crested waves and tests with short-crested waves. The results show that for rock slopes the influence of oblique waves is larger for long-crested waves. Based on the test results a design guideline is provided to account for effects of oblique waves on the stability of rock slopes, armour layers with a double layer of cubes, and armour layers with a single layer of cubes.     

电子罗盘校准方法的设计与实现

电子罗盘在波浪浮标中用于实时提供波浪浮标所处的方位.为解决嵌入了Matlab的电子罗盘校准系统可移植性差、过程繁杂、运行速度慢的问题,设计了新的校准系统.将由14位串行A/D转换器Max194从波高数据采集系统中获得的模拟信号转换成的数字信号接入到微控制器中,微控制器负责控制方位数据的采样频率和A/D转换芯片,再利用串口转换芯片将输出的电平信号转换成RS232信号,由计算机对此进行处理.该系统采用了绘图功能完备、处理速度出众、可移植性好的ProEssentials结合VC++来处理多功能方位测试台采集到的方位数据,并绘制成采样数据比较图和误差图,来直观地展现电子罗盘的误差情况.实验测试表明,该电子罗盘校准系统具有可移植性好、功耗低、稳定性强、运行速度快的特点.该校准系统已经广泛应用于波浪浮标、3m浮标等的生产中.     

基于虚拟仪器技术的海洋移动气象观测系统

介绍一种自主研发的、基于Labview虚拟仪器技术的海洋移动气象观测系统,它可广泛用于各种舰船、海洋移动平台上,能够有效地去除安装平台的移动给风参数测量造成的影响,从而给出真风速、真风向、温度、湿度、气压、安装平台移动轨迹、速度等参数,并能够实时地显示相关参数的随时间的变化趋势,为海洋气象的研究及预报工作提供真实有效的原始数据分析资料。     

Investigation on a damaged ship model sinking into water based on three dimensional SPH method

Damaged ship at sea will be a direct threat to lives and property, and it has a great significance of studying ship's remaining buoyancy, stability, sinking time and other important parameters. The process of a damaged ship sinking into water is a complex motion involving ship hull, inner and outer fluid coupled with waves and many other factors. It is featured by high nonlinearity and hard to establish a precise theoretical model to study. Yet SPH (smoothed particle hydrodynamics) as a meshfree method has a great advantage in solving such problems because of the nature of self-adaptive and Lagrangian. Firstly, the experiments of two scaled ship models with different openings sinking into water are carried out, through the sinking processes of broadside opening and bottom opening models, the conclusion is drawn that although the serious loss of stability of broadside opening model, the sinking time and other parameters are more conducive to rescue after maritime distress. Secondly, the parallel program of three dimensional SPH is developed to simulate the above more complex model, broadside opening model. The coupled process of sloshing is compared with that of experiment, and the results show good agreement with each other which verify the accuracy and feasibility of three dimensional parallel program.