COMPILE-TIME ANALYSIS AND OPTIMIZATION OF EXPLICITLY PARALLEL PROGRAMS*

Data flow analysis has been used by compilers in diverse contexts, from optimization to register allocation. Traditional analysis of sequential programs has centered on scalar variables. More recently, several researchers have investigated analysis of array sections for optimizations on modern architectures. This information has been used to distribute data, optimize data movement and vectorize or parallelize programs. As multiprocessors become more common-place. we believe there will be considerable interest in explicitly parallel programming languages. In this paper, we extend traditional analysis to array section analysis for parallel languages which include additional control and synchronization structures.

We show how to compute array section data flow information, i.e., Communication Sets, for a class of parallel programs. To illustrate its use, we show how this information can be applied in compile-time program partitioning. Information about array accesses can also be used to improve estimates of program execution time used to direct runtime thread scheduling.     

COMPUTING MAXIMUM CLIQUES OF CIRCULAR ARCS IN PARALLEL*

Given a set S of n proper circular arcs, it is required to identify a largest cardinality subset K[S] of S each two of whose members intersect. This paper describes an optimal parallel algorithm to compute K[S]. The algorithm is not based on any previously known sequential solution, and is designed for the CREW PRAM model of computation. It uses 0(n/logn) processors and runs in O(logn) time. An interesting feature of the algorithm is that it transforms the computational geometric problem at hand, to a problem involving computations on 0-1 matrices, and then transforms the latter back into a ray shooting problem in computational geometry.     

三维标量场并行等值面提取与绘制技术

通过研究并行等值面提取与绘制的各项关键技术要点,面向三维标量场的特点,在sort-last并行绘制模式的基础上提出一个三维标量场并行等值面提取与绘制框架.该框架在任务分配时采用静态分配的模式,在等值面提取时采用Marching Tetrahedra等值面提取算法,在并行绘制与场景合成时采用预先构建绘制节点控制模型的混合场景并行绘制合成算法.实验结果表明,该框架能够有效地提高大规模时变三维标量场的等值面提取与绘制效率,满足实际应用的需要.     

基于SOM和引力场聚类的金融数据可视化

平行坐标技术是信息可视化中重要的分析手段,可以实现多维数据在二维空间上的可视化.为了给用户提供一种快捷、方便的金融数据可视化及分析工具,提出一种基于引力场聚类的金融数据可视化方法.首先利用自组织映射(SOM)对初始金融数据进行分类,使每类数据都含有特定的经济意义;然后进行视觉聚类,利用引力场原理对每个类中的折线进行聚拢,对类与类之间进行排斥,再通过设置不透明度以及交互操作等手段对可视化结果进行增强.实验结果表明,该方法可以形成清晰的可视化聚类结果,便于发现数据的变化规律.     

基于CAE的平流式换热器无模成型工艺开发

为实现高效的平流式换热器柔性弯曲制造,避免模具对制品的损伤,采用仿真与试验相结合的方法实现平流式换热器无模弯曲工艺开发.与传统的换热器绕弯成型技术相比,该成型过程无须弯曲模,而是通过控制换热器两端的运动轨迹实现弯曲成型,改变相应的运动轨迹能实现不同的弯曲半径.在成型原理分析的基础上,建立无模弯曲成型过程的有限元仿真模型,并基于仿真结果确定合理的弯曲速度、夹头长度和中性层半径等工艺参数.设计并制造无模弯曲成型试验装置,并研究翅片结构等因素对成型质量的影响.     

基于CPU／GPU异构平台并行优化的研究

CPU／GPU异构系统具有很大的发展潜力,深入研究CPU／GPU异构平台的并行优化,可实现系统整体计算能力的最大化。通过对CPU／GPU任务划分的优化来平衡CPU和GPU的负载,可提高计算资源的利用率,缩短计算任务的执行时间;通过对GPU线程划分的优化,可使GPU获得更高的速度。从而提高系统整体性能。     

意大利RIS-2K探地雷达数据解密及其快速可视化方法

RIS-2K探地雷达数据结构一直不公开,给利用RIS-2K雷达实测数据带来了极大的困难。为此,对RIS-2K探地雷达数据进行了分析与多方测试,基本摸清其数据结构,并编程实现了对RIS-2K探地雷达数据的读取。在常规可视化方法基础上,应用GPU并行技术实现了RIS-2K探地雷达数据快速可视化,有效解决了常规可视化速度慢及界面冻结问题,满足了对实测探地雷达资料进行实时处理解释工作的需要。试验结果证明了该方法的正确性和有效性。     

基于APCNN和BiGRU-Att的单词DGA域名检测方法

为了提高对基于单词的域名生成算法（domain generation algorithm,DGA）生成的恶意域名的检测准确率,提出了一种结合改进的并行卷积神经网络（APCNN）和融合简化注意力机制的双向门控循环单元（BiGRU-Att）的网络模型,该模型能充分学习单词特征、单词之间的组合关系和关键字符信息。实验结果表明,相比Bilbo和CL模型,APCNN-BiGRU-Att模型的分类准确率和F1值更高,表明该模型具有更好的检测效果、多分类效果和稳定性。     

岩石高边坡稳定性并行计算及其应用

在高地震烈度区的岩石高边坡抗震安全评价需要大规模数值计算,而现有商业软件无法有效解决这类复杂的工程问题,因此,本文开发了解决岩石高边坡接触稳定性问题的对等架构并行计算程序,并应用于泰山公司抽水蓄能电站上水库左岸边坡的静力和动力稳定性评价。在数值建模过程中,将岩石边坡滑块体、地质结构面、库水作用以及近场地基作为岩石高边坡动力接触系统。在稳定性安全评价中结合了有限元法、极限平衡法和接触算法,考虑影响高边坡稳定性的多种复杂因素,通过计算特定滑动面的安全系数或安全系数时程,基于现行水电工程水工建筑物抗震设计规范,完成了岩石高边坡稳定性评价,回避了求解复杂岩体材料非线性的问题。计算结果显示,所开发的程序具有较高的并行计算效率,对解决复杂岩石高边坡稳定问题具有显著的优势。     

RAFNet: RGB-D attention feature fusion network for indoor semantic segmentation

Semantic segmentation based on the complementary information from RGB and depth images has recently gained great popularity, but due to the difference between RGB and depth maps, how to effectively use RGB-D information is still a problem. In this paper, we propose a novel RGB-D semantic segmentation network named RAFNet, which can selectively gather features from the RGB and depth information. Specifically, we construct an architecture with three parallel branches and propose several complementary attention modules. This structure enables a fusion branch and we add the Bi-directional Multi-step Propagation (BMP) strategy to it, which can not only retain the feature streams of the original RGB and depth branches but also fully utilize the feature flow of the fusion branch. There are three kinds of complementary attention modules that we have constructed. The RGB-D fusion module can effectively extract important features from the RGB and depth branch streams. The refinement module can reduce the loss of semantic information and the context aggregation module can help propagate and integrate information better. We train and evaluate our model on NYUDv2 and SUN-RGBD datasets, and prove that our model achieves state-of-the-art performances.